relativization in sasak and sumbawa, eastern indonesia*

LANGUAGE AND LINGUISTICS 94865-916 2008 2008-0-009-004-000286-1

Relativization in Sasak and Sumbawa Eastern Indonesia

Masayoshi Shibatani Rice University

On the basis of a detailed study of the relativization phenomena in the dialects of Sasak and Sumbawa in eastern Indonesia this paper shows that the two crucial assumptions made by Keenan amp Comrie on Western Malayo-Polynesian languages are untenable That is the Topic in these languages cannot be reinterpreted as Subject and that Austronesian non-Actor focus constructions cannot be considered passive Both Sasak and Sumbawa just like Standard Indonesian and other Western Malayo- Polynesian and Formosan languages have two privileged but distinct grammatical relations Topic and Subject and the fact that the Topic is most accessible to relativi-zation in these languages falsifies Comrie amp Keenanrsquos claim that ldquoin absolute terms Subjects are the most relativizable of NPrsquosrdquo (Comrie amp Keenan 1979653) The pronoun-retention strategy in Sasak and Sumbawa also falsifies their other universal ldquoAll RC strategies must operate on a continuous segment of the AHrdquo (p661) The paper concludes with the suggestion that relativization in Western Malayo-Polynesian (and Formosan languages and many other non-Austronesian languages) involves nominalized clauses juxtaposed to a head noun in appositive syntagm and this possibility renders moot the question of grammatical relations in relativization Key words Topic Subject Western Malayo-Polynesian languages relativization

strategies nominalization

1 Introduction

Austronesian languages especially the so-called Philippine-type Western Malayo- Polynesian languages have played no small role in typological studies of relative clauses For example Keenan amp Comrie (1977) one of the most celerabrated achievements in modern syntactic typology and their closely related works (Comrie amp Keenan 1979 Keenan 1985 Comrie 1998ch7) almost exclusively rely on the data from Austronesian The research reported here was supported by the National Science Foundation grant BCS-0617198

for the project ldquoAustronesian voice systems an eastern Indonesian perspectiverdquo headed by the author I am grateful to many of the speakers of Sasak and Sumbawa who helped my research on their languages I might note that my language consultants in Lombok include Dr Herman Suheri a Puyung Meno-Meneacute speaker upon whose speech the contributions to Austin (1998)

Masayoshi Shibatani

866

in establishing the higher end of the hierarchy of grammatical relations known as the Accessibility Hierarchy Philippine-type languages are characterized in terms of what is known as the focus morphology that marks in the verb the grammatical role of a nominal element selected as the primary argument of a clause With the crucial assumption that this primary argument is Subject in these languages Keenan amp Comrie observe that these languages obey the ldquoSubjects-onlyrdquo constraint representing languages that relativize only on Subjects and draw the universal ldquoin absolute terms Subjects are the most relativizable of NPrsquosrdquo (Comrie amp Keenan 1979653)

Philippine-type languages where the more faithful reflexes of the Proto Austronesian (PAn) focus opposition are seen are found in Taiwan and Madagascar besides the Philippines The PAn focus system is generally believed to have had three or possibly four categories of AF (Actor-focus) PF (Patient-focus) LF (Location-focus) and IF (Instrumental-focus) eg the indicative neutral series Actor ltumgt Patient -ən Location -an InstrumentalCircumstantial Si- (cf Ross 200233) The four-way PAn focus morphological opposition has been reduced to a three- or two-way system in many of the languages of Indonesia (and elsewhere) where the two-way morphological opposition is seen in Standard Indonesian (Bahasa Indonesia) Javanese Balinese and others in terms of a nasal prefix (for the AF) and a zero prefix or oral form (for the PF) Keenan amp Comriersquos Subjects-only constraint on relativization which Ross (1995730) believes to have obtained in PAn is clearly seen in those languages with the two-way morphological focus opposition The Actor nominal in an AF construction with a nasal verbal prefix and the Patient nominal in a PF construction with a zero prefixmdashboth of which Keenan amp Comrie would consider to be Subjectsmdash are the only directly relativizable arguments in these languages The two-way morphological focus opposition seen in major Indonesian languages as well as in Standard Malay (Bahasa Melayu) is being gradually lost in Sasak and Sumbawa in eastern Indonesia It is an interesting question whether the Subjects-only constraint on relativization has also been lost in the languages that have lost the morpho-logical focus contrast as in many dialects of Sasak and Sumbawa and this is one of the questions addressed in this paper

The paper consists of a descriptive and a theoretical component It offers a detailed description of the relativization phenomena seen in various Sasak dialects and primarily two Sumbawan dialects While typological studies such as Keenan amp Comrie (1977) and others by necessity heavily reply on descriptive grammars of various languages it is true

were almost exclusively based This work will be critically reviewed below My special thanks

go to Nur Ahmadi for his assistance in gathering and interpreting the relevant data I am indebted to Fay Wouk and I Wayan Arka Co-PIs of the NSF project mentioned above who have provided me with comments very helpful in improving this paper I also thank Naonori Nagaya for expertly drawing the three maps in this article


867

that most available grammars spend at most several pages on relativization This study shows that describing the whole picture on relativization phenomena even in a single language requires the space of more than a few pages and that facts crucial in drawing important generalizations are often not available in a short sketch found in many descriptive grammars Another problem in relying on secondary source is that available descriptions may be incomplete or totally incorrect Additionally examination of more than one dialect of a language brings to light interesting variations that will not be available in a study of a single dialect Below we shall demonstrate this to be the case in the currently available descriptions of relativization in Sasak and Sumbawa

There is also a methodological issue involved in describing individual languages While the American structuralist doctorine of analyzing each language in its own right without being influenced by the established grammatical traditions of Europe and else-where or by the structures of other languages is venerable one cannot deny the fact that languages are a historical product showing continuity with their parent languages and similaries with their sister languages Most Indonesian languages for example are Aus-tronesian languages sharing their ancensters with the languages in the Philippines Taiwan Madagascar and others that are more remotely connected genetically The methodological stance taken in this paper is that we profit a great deal by viewing individual languages in a larger perspective that takes into account historical continuity with sister languages We shall demonstrate that in describing Indonesian languages it is useful to keep in mind that the focus constructions and associated phenomena have the history of perhaps more than 6000 years and may constitute an indelible Austronesian trait that may lurk in every grammar In particular we hold Actor-focus (AF) and Patient-focus (PF) constructions as construction types seen across Western Malayo-Polynesian and Formosan languages sharing crucial syntactic properties to be explicated in this study1

On the theoretical front we make the following points First we show that Keenan amp Comriersquos two key assumptions regarding Austronesian languages are untenable That is the primary argument in Philippine-type languages or more broadly Western Malayo- Polynesian and Formosan languages is not Subject and PF and other non-Actor focus constructions are not passive We argue that these Austronesian languages have two distinct

1 Here I have in mind recent studies on Indonesian languages such as Durie (1985) Klamer (1998)

Donohue (1999) and Arka (2003) which would have been much more informative as far as their Austronesian characters are concerned had they taken the stance advocated here My recent fieldwork on a number of Central Malayo-Polynesian including isolating languages of Flores Island in the Nusa Tenggara region of eastern Indonesia does indicate that the Austronesian AFPF structural (not morphological) contrast is far more widespread than has hitherto been assumed Here and below ldquoIndonesian languagesrdquo refers to the Austronesian languages of Indonesia

Masayoshi Shibatani

868

grammatical relations Topic and Subject and that it is Topic that is relevant to relativization phenomena These languages in other words do not fit in with the Keenan-Comrie framework of treating relativization in terms of the hierarchy of grammatical relations such as Subject Direct Object Indirect Object etc

Schachterrsquos (1976) seminal paper on grammatical relations in Tagalog should have been presented by him and understood by others in a much more straightforward manner Namely Schachterrsquos Actor category embracing the S of an intransitive clause and the A of a transitive clause could have been considered Subject in the first place as this grouping of S and A is what constitutes the Subject category in English and others in which this is a well-established grammatical relation Schachterrsquos (and traditional Philippinistsrsquo) Topic could have been recognized as a grammatical relation distinct from Subject Languages possessing two privileged grammatical relations of Topic and Subject are seen in East Asia in Japanese Korean and Chinese where the former two have distinct particles (wa and ga in Japanese) marking a Topic and a Subject A major difference between the Philippine and the East Asian situation is that Topic in the former is a much more grammaticalized category than in the latter in the sense that it is an integral and indispensable argument that plays many syntactic roles associated with Subject in those languages that have not grammaticalized a distinct Topic relation (see Shibatani 1991 on this point) This paper underscores this point delineating crucial distinctions between these two grammatical relations which we claim to obtain in Indonesian (and Malagasy as well as Formosan) languages

Indonesian languages unlike their relatives in the Philippines Taiwan or Madagascar are endowed with passive constructions that are useful in explicating the nature of PF constructions though this advantageous situation has not been fully exploited in past studies of Indonesian languages For example Chung (1976) like many traditional Indo-nesian or Malay grammarians incorrectly analyzes Indonesian PF constructions as passive We show below that PF and passive constructions have distinct grammatical properties Moreover explication of the differences between them requires two separate grammatical relations of Topic and Subject as suggested above

Finally we shall briefly examine the relationship between relativization and nominalization and shed some light on the nature of so-called headless relative clauses with the suggestion that they are better treated as nominalized clauses than relative clauses We conclude this paper with the suggestion that relativization in Indonesian languages (and in many others) involves nominalized clauses juxtaposed with a head noun in appositive syntagm rather than a full clause or sentence whose NP coreferential with the head undergoes movementextraction or deletion The suggested analysis as it turns out renders the question of grammatical relations in relativization rather moot


869

2 Sasak and Sumbawa

Sasak is spoken on Lombok Island and Sumbawa in the western half of Sumbawa Island These two islands lie immediately east of Bali in the province of Nusa Tenggara Barat According to the recent study by Adelaar (2005) the Balinese-Sasak-Sumbawa group together with Malayic and Chamic forms a branch of the newly proposed Malayo- Sumbawan subgroup of the Western Malayo-Polynesian languages From the point of view of genetic grouping and geographic distribution Balinese Sasak and Sumbawa should ideally be studied together (See Maps 1 and 2)

Map 1 Map of Indonesia

Map 2 Map of Nusa Tenggara

This paper concentrates on Sasak and Sumbawa which are likely to be genetically closer to each other than to Balinese While Balinese maintains a two-way morphological focus

Masayoshi Shibatani

870

contrast between AF and PF the morphological focus opposition in transitive clauses has been lost in the majority of Sasak and Sumbawan dialects There is a great deal of dialectal variation in Sasak but the main division we are interested in here is between those that maintain the morphological contrast in transitive clauses and those that have lost it2 The dialects spoken in and around the town of Selong in the east generally appear to maintain the nasalOslash morphological opposition similar to the focus system in other major Indonesian languages including Balinese Three dialects (Selong and Pancor Ngeno-Ngeneacute and Suralaga Ngeto-Ngeteacute) studied to date by the author maintain the morphological focus opposition for most transitive verbs The dialects that have lost the morphological opposition in the transitive use of verbs may still retain nasal forms which may be used as an alternate to an oral form in a transitive construction or may be used as intransitive verbs as illustrated by the Narmada and Sumbawan nasal intransitive use in (4) and (6) below Still speakers of different dialects are likely to vary in terms of which nasal verbs have dropped out from their speech entirely For example my principal consultant of the Bagu Meno-Meneacute dialect has single oral forms bace lsquoto readrsquo and beli lsquoto buyrsquo in both transitive and intransitive use as in (5) below She however retains both oral and nasal forms for bengngebeng lsquoto giversquo and tulisnulis lsquoto writersquo used in the manner of Selong Ngeno-Ngeneacute forms (4) The complex patterns like this are expected in an on-going change in the lexicon As the transition from (4) to (5) suggests the pattern of change at the construction level is that oral forms are gradually spreading across different construction types The locations of the dialects sampled for this study are marked in Map 3 below

2 As shown below dialect divisions with regard to the maintenance or loss of the focus mor-

phology are actually not clear-cut Sasak dialects have been named after their forms for the expression ldquolike this-like thatrdquo Four or five dialects so identified are Ngeno-Ngeneacute Ngeto-Ngeteacute Meno-Meneacute Meriaq-Meriku and Kuto-Kuteacute There is however little linguistic evidence so far supporting this kind of dialect grouping (see Jacq 1998 for the relevant discussion) Indeed the two dialects using the Ngeno-Ngeneacute form that I have studied are different in the aspect crucial to our discussion The Pancor Ngeno-Ngeneacute near Selong in the east maintains the morphological focus opposition while the Narmada Ngeno-Ngeneacute near the provincial capital of Mataram in the west does not with many verbs Because of such variation I identify each dialect with a town name followed by the form for the expression ldquolike this-like thatrdquo as done here It must be noted however that the number of speakers from each town that I have worked with to date is very limited and that much more in-depth work with a larger number of speakers needs to be done for a thorough characterization of each Sasak and Sumbawan dialect


871

Map 3 Map of Lombok Island

(1) Standard Indonesian (Bahasa Indonesia)3 a Saya membaca buku ini (N-AF) I Nread book this lsquoI read this bookrsquo b Buku ini saya baca (Oslash-PF) book this I Oslashread lsquoI read this bookrsquo (2) Balinese a Tiang mamaca buku=ne (N-AF) I Nread book=DEF lsquoI read the bookrsquo

3 Abbreviations used in this paper are as follows empty place where a relevant item is missing or a

zero morpheme 1 1st person 2 2nd person 3 3rd person A transitive agentive participant AF actor focus construction APPL applicative ART article CONS consequence COP copula DEF definite DO direct object FUT future GEN genitive IO indirect object LIN linker LOC locative LOW low register MASC masculine N nasal AF prefix NMZ nominalizernominalization marker NOM nominative NON-TOP non-topic OBJ object OBL oblique object OCOMP object of comparison P transitive patientive participant PASS passive PAST past tense PERF perfective PF patient focus construction PROG progressive REL relativizerrelative clause marker S single participant of an intransitive clause SG singular SU subject SUB subject TOP topic

Masayoshi Shibatani

872

b Buku=ne baca tiang (Oslash-PF) book=DEF read I lsquoI read the bookrsquo (3) Selong Ngeno-Ngeneacute Sasak a Eku mbace buku ino (N-AF) I Nread book this lsquoI read that bookrsquo b Buku ino ku=bace (Oslash-PF) book this 1=read lsquoI read that bookrsquo (4) Narmada Ngeno-Ngeneacute Sasak a Aku jengke-ng=ku bacembace buku=ni (Oslash-N-Transitive) I PROG-LIN=1 OslashreadNread book=this lsquoI am reading this bookrsquo b Aku jengke-ng=ku mbace (N-Intransitive) I PROG-LIN=1 Nread lsquoI am readingrsquo (5) Bagu Meno-Meneacute Sasak a Aku nyenke-ng=k bace buku=ne (Oslash-Transitive) I PROG-LIN=1 read book=this lsquoI am reading this bookrsquo b Aku nyenke-ng=k bace (Oslash-Intransitive) I PROG-LIN=1 read lsquoI am readingrsquo (6) Sumbawa Sumbawa Taliwang a Buku=sa ku=bacamaca (N-Oslash-Transitive) book=this 1=readread lsquoI read this bookrsquo b Sarsquo muntu ku=bacamaca (N-Oslash-Intransitive) now 1=readread lsquoI am reading nowrsquo Sumbawa Besar a Aku baca buku ta (Oslash-Transitive) I read book this lsquoI read this bookrsquo b Ta muntu ku=bacamaca (Oslash-N-Intransitive baca preferred) now 1=readread lsquoI am reading nowrsquo


873

3 Subject and Object in Sasak and Sumbawa

Subjects and Objects are both robust syntactic relations in Sasak and Sumbawa The clearest indications for these categories come from patterns of cliticization Whereas the pattern of cliticization of Subjects to main verbs varies somewhat from one dialect to another cliticization of Subjects to auxiliaries is more consistent across Sasak dialects All of them allow a Subject to cliticize to an auxiliary although the cliticization may be optional or strongly disfavored by some speakers because of redundancy when a full pronoun occurs Only Subjects can cliticize to auxiliaries

(7) Puyung Meno-Meneacute Intransitive Subjects a (Aku) mu=k lalo jok peken4 I PAST=1 go to market lsquoI went to the marketrsquo b Mu=m lalo jok peken PAST=2 go to market lsquoYou went to the marketrsquo c Inaq mu=n lalo jok peken mother PAST=3 go to market lsquoMother went to the marketrsquo d Mu=n lalo jok peken PAST=3 go to market lsquoShe went to the marketrsquo Transitive Subjects e Mu=k empuk Ali PAST=1 hit Ali lsquoI hit Alirsquo f Inaq mu=n empuk Ali mother PAST=3 hit Ali lsquoMother hit Alirsquo g Mun=n empuk Ali PAST=3 hit Ali lsquoShe hit Alirsquo

4 Precise characterizations of the auxiliaries glossed as PROG PAST FUT and PERF here and elsewhere

require further studies

Masayoshi Shibatani

874

The phenomenon of cliticization seen here is Subject-centered and includes not only intransitive Srsquos and transitive Arsquos but also the P of a passive clause while it excludes the A of a passive as in exactly the same way the English Subject category does

(8) Puyung Meno-Meneacute Passive Subjects

a (Aku) wah=k te-empuk isiq Ali I PERF=1 PASS-hit by Ali lsquoI have been hit by Alirsquo b Te-empuk=m isiq Ali PASS-hit=2 by Ali lsquoYou were hit by Alirsquo c Te-empuk=n isiq Ali PASS-hit=3 by Ali lsquoShe was hit by Alirsquo

Some dialects eg Puyung Meno-Meneacute have a richer and fuller set of Object enclitics than some others with a more impoverished system like Selong and Pancor Ngeno-Ngeneacute which have only a third person Object enclitic Notice in the following examples from Puyung Meno-Meneacute that Object clitic doubling is also possible in this dialect There is a clear boundary between an Object argument and an adjunct such that the latter can cliticize to the verb only after becoming an Object via applicativization as in (9h) below

(9) Puyung Meno-Meneacute a Ali empuk=k Ali hit=1 lsquoAli hit mersquo b Mu=k empuk-e=m PAST=1 hit-LIN=2 lsquoI hit yoursquo c Aku wah=k gitaq=n Ali I PERF=1 see=3 Ali lsquoI have seen Alirsquo d Mu=k beng inaq kelambi PAST=1 give mother shirt lsquoI gave mother a shirtrsquo


875

e Mu=k beng=n kelambi PAST=1 give=3 shirt lsquoI gave himher a shirtrsquo f Yaq=k jauq buku=no jok Siti FUT=1 take book=the to Siti lsquoI will take the book to Sitirsquo g Yaq=k jauq-ang Siti buku=no (Applicative) FUT=1 take-APPL Siti book=the lsquoI will take the book to Sitirsquo h Yaq=k jauq-ang=n buku=no FUT=1 take-APPL=3 book=the lsquoI will take the book to himherrsquo

The Sumbawan cliticization pattern appears similar to the Sasak pattern except that there is no third person clitic form (cf (10eprime) below) Note that subject markers are proclitics whereas object markers are enclitics

(10) Sumbawa Besar a Ka=ku=naneacutengrsquo aku (Shiohara 200657) PERF=1SGLOWAFFIX=takeashower 1SGLOW lsquoI have taken a showerrsquo b Ka=ku=pukil=mu PERF=1SG=hit=2SG lsquoIrsquove hit yoursquo c hellipada tau ka-ingo=ku teris angkat=kuhellip (Wouk 2002302) exist person PERF-see=1SG then lift=1SG lsquohellipthere was someone who saw me then lifted mersquo d Surat=nan mu=tulis letter=that 2SG=write lsquoYou wrote the letterrsquo e Nya ka=alo ko amat she PERF=go to market lsquoShe went to the marketrsquo eprime ka=nya=alo ko amat (cf (10a-b) above) PERF=3SG=go to market lsquoShe went to the marketrsquo f Aku na=i=pukul=ku ning Ali (Sumbawa Taliwang passive) I FUT=PASS=hit=1SG by Ali lsquoI will be hit by Alirsquo

Masayoshi Shibatani

876

Besides the cliticization of Subjects and Objects there are a number of phenomena sensitive to the relational categories of Subject and Object in these languages Postponing our discussion of the phenomena in which the Subject category plays a crucial role until a later section let us examine the robustness of the Object category in Sasak While basic Objects can be passivized objects of prepositions cannot be In order for such Objects to be passivized they must first become Objects via applicativizationmdashin the same way that these prepositional objects must first become applicative Objects before cliticizing to verbs As examples (11b-c) below show both Objects of a Double Object Construction can be passivized in Sasak

(11) Pancor Ngeno-Ngeneacute a Loq Ali ngembeng kanak=no buku ino (AF construction) ART Ali Ngive child=that book that lsquoAli gave the child that bookrsquo b Kanak=no te-embeng buku ino siq loq Ali (Passive) child=that PASS-give book that by ART Ali lsquoThe child was given that book by Alirsquo c Buku ino te-embeng kanak=no siq loq Ali (Passive) book that PASS-given child=that by ART Ali lsquoThat book was given (to) the child by Alirsquo d Loq Ali ngirim surat timpaq batur=ku (AF) ART Ali Nsend letter to friend=1 lsquoAli sent a letter to my friendrsquo e Loq Ali ngirim-ang batur=ku surat (Applicative AF) ART Ali Nsend-APPL friend=1 letter lsquoAli sent my friend a letterrsquo f Batur=ku te-kirim-ang surat siq loq Ali (Passive) friend=1 PASS-send-APPL letter by ART Ali lsquoMy friend was sent a letter to by Alirsquo

Finally all these Objects (basic and derived applicative Objects in Sasak and basic

Objects in SumbawamdashSumbawa appears to have no applicatives) but not the Objects of prepositions can be Patient focused ie become the Topics of PF constructions Remember that Pancor Ngeno-Ngeneacute has a nasaloral morphological contrast indicating the AFPF structural contrastmdashthe nasal AF construction has an A Topic and the oral PF construction a P Topic


877

(12) Pancor Ngeno-Ngeneacute PF constructions a Kanak=no embeng=na buku ino siq loq Ali5 (cf AF form 11a) child=that Oslashgive=3 book that by ART Ali lsquoAli gave the child that bookrsquo b Buku ino embeng=na kanak=no siq loq Ali (cf AF form 11a) book that Oslashgive=3 child=that by ART Ali lsquoAli gave the child that bookrsquo c Batur=ku kirim-ang=na surat siq loq Ali (cf AF form 11e) friend=1 Oslashsend-APPL=3 letter by ART Ali lsquoAli sent my friend a letterrsquo

Notice that in these PF constructions it is still the agentive Subjectsmdashnot the

Topicsmdashthat cliticize to auxiliaries despite the fact that these postposed agentive Subjects just like the passive agents are preceded by a preposition siq being translated as lsquobyrsquo here Compare the following two sets of sentences6

(13) Pancor Ngeno-Ngeneacute a Loq Alii wah=nei ngirim=ang oku surat (AF) ART Ali PERF=3 Nsend=APPL I letter lsquoAli sent me a letterrsquo aprime Oku wah=nei kirim-ang surat siq loq Alii (PF) I PERF=3 Oslashsend-APPL letter by ART Ali lsquoAli sent me a letterrsquo b Okui wah=kui ngirim-ang loq Ali surat (AF) I PERF=1 Nsend-APPL ART Ali letter lsquoI sent Ali a letterrsquo

5 The form siq and its equivalent isiq in other dialects found in these constructions marking the

agentive nominal is apparently the same form used in marking the agent of a passive clause These two constructions however are different in that passive clauses have a passive morphology in the verbmdashsee the passive forms in (11) The two also differ structurally in a crucial way as shown in the subsequent sections

6 Translation of PF forms into English is problematic Many Austronesian specialists translate PF forms into English passive sentences but they are not really like the English passive in that the postposed agents of PF constructions unlike the passive agents in English and in these Indonesian languages retian high discourse relevance as reflected in their syntactic status as Subject Voice functions of PF constructions must be explicated in assessing the total grammatical properties of PF constructions Awaiting further studies English glosses for PF constructions are given in the active form

Masayoshi Shibatani

878

bprime Loq Ali wah=ku kirim-ang surat (PF) ART Ali PERF=1 Oslashsend-APPL letter lsquoI sent Ali a letterrsquo

4 Relativization

Having established the grammatical relations Subject and Object in Sasak and Sumbawa we are now in a position to examine the relativization patterns in these languages The two works most immediately relevant to our discussions are Eades (1998) and Shiohara (2000) The former describes Puyung Meno-Meneacute Sasak relativization in terms of the Accessibility Hierarchy (given below) and the latter Sumbawa Besar relativization in reference to the Dixonian S A and P Both these languages have lost the morphological focus contrast in the majority of transitive verbs with existing nasal verbs typically occurring as intransitive verbs

(14) Accessibility Hierarchy (Keenan amp Comrie 197766) SU gt DO gt IO gt OBL gt GEN gt OCOMP

After examining the relative clauses contained in examples like the ones below Eades (1998128-129) concludes his work on Puyung Meno-Meneacute relativization by noting that ldquo[a]ll of the categories in Keenan amp Comriersquos accessibility hierarchy are accessible to relativization in Sasak except for the Object of comparisonrdquo and that ldquo[t]he process of relativization in Sasak is quite different from that in other Western-Austronesian languages many of which can only relativize a syntactic Subject and require some other form of promotion of the non-Subject to Subject in order for it to be accessible to relativizationrdquo7

(15) Puyung Meno-Meneacute a Mu=k gitaq acong [saq kakoq=em wiq]=no (SU Eades 1998124) PAST=1 see dog REL bite=2 yesterday=that lsquoI saw the dog that bit you yesterdayrsquo b inaq kelor sebie [saq mu=n beli wiq] (DO Eades 1998124) mother eat chili REL PAST=3 buy yesterday lsquoMother ate the chili that she bought yesterdayrsquo

7 Apparently Peter Austin Yusuf Eadesrsquo teacher has a paper entitled ldquoRelativization in Sasak

eastern Indonesiardquo which he read at the Australian Linguistic Society conference in Brisbane in 1998 and in which he talked about the relativization phenomena in different Sasak dialects To the best of my knowledge this paper has not been published It was not available to me


879

c dengan mame [saq mu=k beacuteng nasiq]=no (IO Eades 1998125) person male REL PAST=1 give rice =that tokol leacuteq meacuteje sit at table lsquoThe man who I gave the rice to sits at the tablersquo d tunjang [saq isiq=n empuk acong isiq (OBL Eades 1998127) stick REL by=3 hit dog by kanak]=no leacuteq atas meacuteja child =that at top table lsquoThe stick which the child hit the dog with is on the tablersquo e mu=k gitaq dengan [saq acong=en (GEN Eades 1998128) PAST=1 see person REL dog=3 kakoq Ali]=no bite Ali=the lsquoI saw the man whose dog bit Alirsquo

Eades believes that (15b) is a case of Direct Object relativization (15c) a case of

Indirect Object relativization (15d) a case of Oblique relativization and (15e) a case of Genitive relativization Except for the Subject relativization case (15a) above all the analyses proposed by Eades in (15b-e) are problematic Indeed further investigation shows that it is not the case that Objects and Oblique objects are straightforwardly relativizable as indicated by the ungrammatical forms below

(16) Puyung Meno-Meneacute a inaq mu=n beli sebie wiq (Eades 1988123) mother PAST=3 buy chili yesterday lsquoMother bought the chili yesterdayrsquo b Anak=no kelor sabie [saq inaq mu=n beli Oslash wiq] child=that eat chili REL mother PAST=3 buy yesterday lsquoThat child ate the chili which mother bought yesterdayrsquo

(16a) taken from Eades is the transitive clause that he believes to be the basis for (15b) However (16b) shows that Object relativization is not acceptable in contrast with (15b) Note that in (16b) unlike in (15b) the actor is doubly marked by a full noun and a clitic on the auxiliary rather than just by a clitic

(17a) below illustrates a typical transitive clause with an oblique in this case an instrumental NP marked by the preposition isiq As shown in (17b) the oblique Object cannot be relativized in a straightforward manner in contrast with (15d) which actually

Masayoshi Shibatani

880

is related to a more elaborate construction (see (55c) below) than to the simple transitive construction in (17a)

(17) Puyung Meno-Meneacute a Anak=no empuk acong isiq tunjang child=DEF hit dog by stick lsquoThe child hit the dog with a stickrsquo b Tunjang [saq anak=no empuk acong (isiq) Oslash]=no leacuteq atas meacuteja stick REL child=that hit dog by =the on top table lsquoThe stick with which the child hit the dog is on the tablersquo What is identified by Eades as an Indirect Object in (15c) is actually the Object of a

double object construction seen earlier (see (11)) Neither of the double objects of such a construction can be relativized straightforwardly as shown by examples (18b-c) below Again note that in the ungrammatical examples (18b-c) A within the relative clause is doubly marked as a full noun and a clitic while in (15c) A is only represented by a clitic

(18) Puyung Meno-Meneacute a Ali mu=n ebeacuteng dengan mame=no nasiq Ali PAST=3 give person male=that rice lsquoAli gave the man ricersquo b Dengan mame [saq Ali mu=n ebeacuteng Oslash nasiq]=no batur=k person male REL Ali PAST=3 give rice=that friend=1 lsquoThe person whom Ali gave rice is my friendrsquo c Nasiq [saq Ali mu=n ebeacuteng dengan mame Oslash]=no poteacuteq rice REL Ali PAST=3 give person male=that white lsquoThe rice that Ali gave to the man was whitersquo

Postponing our discussion of the relativization of the Genitive for now we turn to

Shiohararsquos (2000) treatment of the Sumbawan relativization which has problems similar to Eadesrsquo Based on the following kind of examples Shiohara implies that S A and P roles are all relativizable

(19) Sumbawa Besar a tentara=nan datang kalis=anosiup (Shiohara 200087) army=that come from=east lsquoThe army came from the eastrsquo


881

b tentara [adeacute=datang kalis=anosiup] (S-rel Shiohara 200087) army REL=come from=east ta=basingin pasukanan Bala Cucuk this=be named troop troop front line lsquoThe army which comes from the east is named lsquoBala cucukrsquo (20) Sumbawa Besar a ya=beri togravedeacute=nan ling=tau=lokaq (Shiohara 200087) 3=like child=that by=person=old lsquoThe parents love the childrsquo b tau=lokaq [deacute=beri=togravedeacute=nan] (A-rel Shiohara 200088) person=old REL=love=child=that lsquothe parents who love their childrenrsquo c togravedeacute [adeacute ya=beri ling=tau=lokaq] (P-rel Shiohara 200088) child REL 3=like by=person=old lsquothe child who is loved by hisher parentsrsquo

Shiohara (2000) however leaves out the important fact that an A in clauses such as

(20a) repeated as (21a) below and a straightforward Object P in sentences such as the one in (22a) cannot be relativized

(21) Sumbawa Besar a ya=beri togravedeacute=nan ling=tau=lokaq 3=like child=that by=person=old lsquoThe parents love the childrsquo b tau=lokaq [adeacute ya=beri togravedeacute=nan (ling) Oslash] person=old REL 3=like child=that by lsquothe parents who love the childrsquo (22) Sumbawa Besar a tau=lokaq beri togravedeacute=nan person=old love child=that lsquoThe parents love the childrsquo b togravedeacute [adeacute tau=lokaq beri Oslash] (cf (20c) above) child REL person=old love lsquothe child whom the parents loversquo

Thus contrary to the explicit claim made by Eades (1998) and to what is implied in

the description by Shiohara (2000) both Sasak and Sumbawa place severe restrictions on relative clause formation Eadesrsquo problem was that he failed to identify correctly the

Masayoshi Shibatani

882

structures underlying his relative clauses whereas Shiohara did not examine all logically possible structures to see whether or not they all yield grammatical relative clauses Upon closer examination of what is and what is not relativizable in these languages it becomes clear that the relativization process in these languages is not very different from that found in other Austronesian languages with the focus morphology

5 Focus constructions without focus morphology

It is a curious fact that neither Austin (2000) nor any of his fellow contributors to the two volumes on Sasak he edited (Austin 1998 2000) has recognized the significance of the distinction between two types of transitive constructions and its relevance to relativization and related phenomena They all clearly acknowledge the existence of two types of transitive constructions as shown below

(23) Puyung Meno-Meneacute (Musgrave 199892) a inaq mu=n kelor sebie odaq mother PAST=3 eat chili green lsquoMother ate green chilirsquo b mu=n kelor sebie odaq isiq inaq PAST=3 eat chili green by mother lsquoMother ate green chilirsquo (24) Puyung Meno-Meneacute (Kroon 1998111) a Herman wa=n ebeacuteng inaq klambi Herman PERF=3 give mother shirt lsquoHerman has given mother a shirtrsquo b Wa=n ebeacuteng inaq klambi isiq Herman PERF=3 give mother shirt by Herman lsquoHerman has given mother a shirtrsquo


883

(25) Mataram (Austin 20006)8 a Aku mbeli baleacute I Nbuy house lsquoI buy a housersquo b Baleacute beli isiq loacuteq Ali house buy by ART Ali lsquoAli bought a housersquo (26) Sumbawa Besar (Shiohara 2006142 143) a Aku ya=kakanrsquo tepoacuteng=ta 1SGLOW CONS=eat cake=this lsquoI will eat this cakersquo b tepoacuteng=ta ya=ku=kakanrsquo leacuteng aku cake=this CONS=1SGLOW=eat by 1SGLOW lsquoI will eat this cakersquo

8 In fairness to Austin we must point out that Austin (2000) does recognize the importance of the

nasaloral contrast in Ngeno-Ngeneacute verbs and notes that this dialect ldquoresembles Balinese hellip in picking out the Agent for a two- or three-place zero verb for special treatment Topicalisation question formation relative clause formation and purpose clause construction in Mataram and Selong Sasak are not possible when the pivot is a zero verb Agentrdquo (14) But then he goes on to say that ldquo[i]n the Menoacute-Meneacute varieties [which have lost the nasaloral opposition in most transitive verbs] this restriction does not apply and any argument of a verb may be directly questionedrdquo (16) and that ldquo[i]n Menoacute-Meneacute Sasak [relativizations] a contrast is made between arguments of verbsand non-argumentshelliprdquo (17) It appears that Austin recognizes the AFPF structural contrast for Ngeno-Ngeneacute dialects which preserve the nasaloral morphological focus contrast in the verb but not for other dialects that have lost the morphological focus contrast Our point below is that the AFPF structural contrast obtains even in those languages that have lost the morphological focus contrast and that it is not true contrary to the claims in the quotes above that in the Meno-Meneacute dialects ldquoany argument of a verb may be directly questionedrdquo or that the only contrast these dialects make for the relativization purpose is between arguments and non-arguments of verbs As for the former claim consider the form Sai kamu empuk (who you hit) lsquoWhom did you hitrsquo based on the AF construction Kamu empuk Ali (you hit Ali) lsquoYou hit Alirsquo This question form which directly questions the Object of the verb empuk lsquohitrsquo was rejected by Dr Herman Suheri Austinrsquos principal consultant for the Puyung Meno-Meneacute dialect The correct form for this question would be Sai mu=m empuk (who PAST=2 hit) lsquoWhom did you hitrsquo which is based on such PF constructions as Mu=m empuk Ali (PAST=2 hit Ali) or Ali mu=m empuk (Ali PAST=2 hit) lsquoYou hit Alirsquo Similarly the form Ape kanak=no wah=n kaken (what child=that PERF=3 eat) lsquoWhat has the child eatenrsquo which directly questions the Object of kaken lsquoeatrsquo was rejected by another Puyung Meno-Meneacute speaking consultant The correct form would be the one based on a PF construction namely Ape wah=n kaken kanak=no (what PERF=3 eat child=that) lsquoWhat has the child eatenrsquo Notice that in these the verb forms remain the same whether they occur in an AF or a PF construction

Masayoshi Shibatani

884

The (b) constructions above have been called ldquothe isiq constructionrdquo by Kroon (1998) and ldquopostposed agent constructionsrdquo by Musgrave (2000) These constructions strike Kroon (1998105) as ldquoa peculiar pattern that distinguishes the Sasak language from all other Western Malayo-Polynesian languagesrdquo The fact that the neighboring Western Malayo-Polynesian language Sumbawa shows the parallel pattern of transitive alternation as in (26) aside a glance at Austinrsquos examples in (25) from Matarammdasha Ngeno-Ngeneacute variety I believemdashclearly indicates that what we observe here is the AFPF structural contrast seen in Formosan languages and across many Western Malayo-Polynesian lan-guages including Standard Indonesian and Balinese Notice that (25a) has a nasal verb form and (25b) the oral version Observe the pattern of correspondence between the sets of AF and PF constructions here and those given for Standard Indonesian Balinese and Selong Ngeno-Ngeneacute in (1)-(3) earlier In other words what we are observing here is the AFPF structural contrast without focus morphologymdashexcept for (25) where we see the widespread morphological NOslash contrast preserved That is in the (a) forms in (23)-(26) above the agent nominals function as Topics while in (b) forms there it is the patientgoal nominals that have been selected as Topics It is this pattern of Topic selection that qualifies these constructions as AF or PF despite their lack of focus morphology which signals such Topic choice in other Western Malayo-Polynesian and Formosan languages

Eadesrsquo findings are thus now easily explained His DO and IO relativizations (15b-c) are actually cases of relativization of the Topic of a PF clause with a morphologically unmarked verb Puyung Meno-Meneacute behaves in a manner exactly parallel to dialects of Sasak that have retained the oralnasal morphological contrast in transitive verbs

It appears that across Sasak dialects and in Sumbawa PF constructions are preferred over their AF counterparts especially in a narrative or reporting of an event This could simply be mirroring the Austronesian preference for PF constructions AF constructions however are fully functional in these languages and could be used in answering questions like ldquoWho ate the chilirdquo or ldquoWho bought the houserdquo9

9 Austin (20006) tells us that in the Praya dialectmdasha variety of Meno-Meneacutemdashldquo[w]hen the Agent

NP is inanimate it must be expressed in an isiq phrase after the verb helliprdquo and gives the following example in (a) He goes on to say that ldquo[w]hen the agent is third person animate and the Patient is also third person hellip then the isiq construction must be used helliprdquo and gives example (b)

(a) Mu=n leacuteoacutep kalambi=noacute isiq angin foc=3 blow shirt=that by wind lsquoThe wind blew the shirt awayrsquo

(b) Mu=n jeloacuteq kelambi= noacute isiq inaq foc=3 dryinsun shirt=that by mother lsquoMother dried the shirt in the sunrsquo All the consultants that I worked with including Dr Herman Suheri Austinrsquos principal consultant of the dialect of Puyung next to the town of Praya report that while these PF constructions are


885

More crucial to our discussion of the relativization phenomena in Sasak and Sumbawa is the very fact that only the A nominals in the (a)-type sentence in (23)-(26) and only the P nominals in the (b)-type sentence in (23)-(26) are relativizable and this is what was shown in (16) (18) (21) and (22) above Just to make the same point if we take (23a) and (23b) above and try to relativize the P nominal in the former and the A nominal in the latter ungrammatical forms result as shown by the (c) forms below

(27) Puyung Meno-Meneacute a (=23a) inaq mu=n kelor sebie odaq (AF) mother PAST=3 eat chili green lsquoMother ate green chilirsquo

b dengan nine [saq Oslash kelor sebie odaq]=no (Topic A relativized) person female REL eat chili green=that inaq=k mother=1 lsquoThe woman who ate green chili is my motherrsquo c Sebie odaq [saq inaq mu=n (Non-Topic P relativized) chili green REL mother PAST=3 kelor Oslash] besar eat big lsquoThe green chili which mother ate was bigrsquo

preferred in reporting the events the following AF constructions are perfectly grammatical and would answer questions such as ldquoWhat did the wind do to the shirtrdquo or ldquoWhat did mother dordquo

(aprime) Angin mu=n leacuteoacutep kalambi=noacute wind PAST=3 blow shirt=that lsquoThe wind blew the shirt awayrsquo

(bprime) inaq mu=n jeloacuteq kelambi=noacute mother PAST=3 dryinsun shirt=that lsquoMother dried the shirt in the sunrsquo Similarly Musgrave (200050) reports that ldquowhen the actor is not a speech act participant hellip and

the undergoer is then only a passive clause is possiblerdquo and gives the following example in (c) from Puyung Meno-Meneacute

(c) Aku te-gitaq isiq Ali I PASS-see by Ali lsquoI was seen by Alirsquo

(cprime) Ali gitaq aku Ali see I lsquoAli sees mersquo

Again AF form (cprime) which Musgrave marks as ungrammatical was found perfectly grammatical by Dr Herman Suheri Musgraversquos chief consultant of Puyung Meno-Meneacute who told me that such a form would be needed in answering the question ldquoWho sees yourdquo

Masayoshi Shibatani

886

(28) Puyung Meno-Meneacute a (=24b) Mu=n kelor sebie odaq isiq inaq (PF) PAST=3 eat chili green by mother lsquoMother ate green chilirsquo b Sebie odaq [saq mu=n kelor Oslash isiq (Topic P relativized) chili green REL PAST=3 eat by inaq] besar mother big lsquoThe green chili which mother ate was bigrsquo c dengan nine [saq mu=n kelor (Non-Topic A relativized) person female REL PAST=3 eat sebie odaq (isiq) Oslash]=no inaq=ku chili green by=that mother=1 lsquoThe woman who ate green chili is my motherrsquo

The above pattern parallels exactly the ones observed in those languages where the AFPF structural contrast is expressed morphologically as in Standard Indonesian Balinese and SalongPancor Ngeno-Ngeneacute Sasak From nasal AF constructions only the A Topic can be relativized and from oral PF constructions only the P Topic can be relativized In the interest of space only the parallel Standard Indonesian and Pancor Ngeno-Ngeneacute Sasak patterns are shown below

(29) Standard Indonesian a Sayaa membeli rumah ini (N-AF) I Nbuy house this lsquoI bought this housersquo aprime Orang [yang Oslash membeli rumah] (A-Topic relativized) person REL Nbuy house itu teman saya that friend I lsquoThe person who bought the house is my friendrsquo aPrime Rumah [yang saya membeli Oslash] (Non-Topic P relativized) house REL I Nbuy itu baru that new lsquoThe house which I bought is newrsquo b Rumah ini saya beli (Oslash-PF) house this I Oslashbuy lsquoI bought this housersquo


887

bprime rumah [yang Oslash saya beli] itu baru (P-Topic relativized) house REL I Oslashbuy that new lsquoThe house which I bought is newrsquo bPrime Saya [yang rumah ini Oslash beli] pintar (Non-Topic A relativized) I REL house this Oslashbuy smart lsquoI who bought this house am smartrsquo (30) Pancor Ngeno-Ngeneacute a Dengan mame=no mbeli baleacute ino (N-AF) person male=that Nbuy house that lsquoThe man bought that housersquo aprime Dengan mame [si Oslash mbeli baleacute] ino (A-Topic relativized) person male REL Nbuy house that batur=ku friend-1 lsquoThe man who bought a house is my friendrsquo aPrime baleacute [si degnan mame=no mbeli Oslash] (Non-Topic P relatived) house REL person male=that Nbuy ino baru that new lsquoThe house which the man bought is newrsquo b Baleacute ino beli=ne siq dengan mame=no (Oslash-PF) house that Oslashbuy=3 by person male=that lsquoThe man bought that housersquo

bprime Baleacute [si Oslash beli=ne siq dengan mame] (P-Topic relativized) house REL Oslashbuy=3 by person male ino baru that new lsquoThe house which the man bought is newrsquo bPrime Dengan mame [si baleacute ino (Non-Topic A relativized) person male REL house that beli=ne (siq) Oslash] ino batur=ku Oslashbuy=3 by that friend=1 lsquoThe man who bought that house is my friendrsquo

These parallel patterns conclusively show that both Sasak and Sumbawan dialects that have lost the morphological focus contrast in the majority of transitive verbs still maintain the AFPF structural contrast seen in other Austronesian languages with the focus morphology More significantly for our present discussion the relativization process in both these types of languages still obeys the Austronesian syntactic constraint that only

Masayoshi Shibatani

888

Topics can be relativized despite their loss of the morphological focus contrast If Ross (1995) is correct about his conjecture that Proto Austronesian had this constraint the constraint has a history of more than 6000 years It makes sense that a constraint of a long history would persist even among those languages that have lost or that are losing the focus morphology in very recent years as appears to be the case with the Sasak and Sumbawan dialects

6 Subject or Topic

The conclusion reached above that the Topic relation is crucially involved in the relativization process in Sasak and Sumbawa is clearly problematic to the approach to relativization based on the Accessibility Hierarchy given in (14) which in turn is defined in terms of the grammatical relations Subject Object and others There is thus good reason for Keenan amp Comrie to consider Austronesian Topics to be Subjects We do not want however to make this move for Sasak and possibly Sumbawa for several reasons

First as demonstrated above and below there are robust Subject and Object categories in addition to the Topic category in Sasak and Sumbawa If the latter were to be considered Subject we would need a separate relational category for the group of nominals that control cliticization and other phenomena (see below) But the problem is that this group of nominalsmdashS A and P of a passivemdashis exactly what constitutes Subjects in those languages eg English where Subject is a well-established syntactic category The basic alignment patterns of participant roles (A and P) SubjectObject and Topic in the three relevant constructions in Sasak (and Sumbawa) are as follows regardless of the presence or absence of the focus morphological contrast

(31) Puyung Meno-Meneacute a AF-construction Ali bace buku=ni (A=SUB=TOP P=OBJ) Ali read book=this lsquoAli reads this bookrsquo b PF-construction Buku=ni mu=n bace siq Ali (A=SUB P=OBJ=TOP) book=this PAST=3 read by Ali lsquoAli reads this bookrsquo c Passive construction Buku=ni te-bace siq Ali (A=OBL P=SUB=TOP) book=this pass-read by Ali lsquoThis book is read by Alirsquo


889

Notice that there are two types of postposed agent nominals in these constructions In the PF form siq marks a postposed agentive Subject which triggers Subject cliticization as in the Sasak (b) form above In passive (c) form on the other hand the preposition siq marks the agentive Oblique nominal which does not trigger Subject cliticization In a passive it is the patientive Subject (Topic) that triggers Subject cliticization Compare the postposed agentive Subject of a PF construction and the agentive oblique of a passive below

(32) Puyung Meno-Meneacute a Mu=n gitaq=k siq nie (PF) PAST=3 see=1 by she lsquoShe sees mersquo b Kamu wah=em te-empuk siq Ali (Passive) you PERF=2 PASS-hit by Ali lsquoYou were hit by Alirsquo Secondly there are several phenomena besides relativization which require a clear

distinction between the Topic and the Subject category As in many languages Sasak makes a distinction between two Equi-type control phenomena The main predicate mele lsquowantrsquo which permits an ldquounlike-Subjectrdquo in the complement clause allows both AF and PF complement clauses and the control pattern is based on the Topic category On the other hand main predicates such as cobaq lsquotryrsquo and suruq lsquoorderrsquo which require a ldquolike-Subjectrdquo in the complement clause coreferential with either the main Subject or Object allow only AF and passive complements and here the control pattern operates in terms of the Subject=Topic alignment10 Let us first look at the former case with mele lsquowantrsquo as a main predicate

(33) Selong Ngeno-Ngeneacute a Mele-ng=ku [anta ngiduk Siti] (AF complement) want-LIN=1 you Nkiss Siti lsquoI want you to kiss Sitirsquo b Mele-ng=ku [Siti meq=iduk] (PF complement) want-LIN=1 Siti 2=Oslashkiss lsquoI want you to kiss Sitirsquo

10 A small minority of speakers fail to make this contrast between the equivalents of lsquowantrsquo and

lsquotryrsquo Also beware of the fact that realis PF forms cannot be embedded under these predicates

Masayoshi Shibatani

890

When this construction contains an ldquoEqui nominalrdquo in either A or P position of a complement clause AF PF and passive clauses may occur provided that the gaps (marked Oslash in the examples below) controlled by the main-clause Subject bear the Topic relation within the complement clause In the examples below the relevant full complement clauses are separately provided as (aprime) (bprime) etc for ease of comprehension

(34) Selong Ngeno-Ngeneacute a Mele-ng=ku [Oslash ngiduk Siti] (Oslash=A=SUB=TOP) want-LIN=1 Nkiss Siti lsquoI want to kiss Sitirsquo aprime [Eku ngiduk Siti] (AF) I Nkiss Siti lsquoI kiss Sitirsquo b Mele-ng=ku [ne=iduk Oslash isiq Siti] (Oslash=P=OBJ=TOP) want-LIN=1 3=Oslashkiss by Siti lsquoI want Siti to kiss (me)rsquo bprime [ne=iduk eku isiq Siti] (PF) 3=Oslashkiss I by Siti lsquoSiti kisses mersquo c Mele-ng=ku [Oslash te=iduk isiq Siti] (Oslash=P=SUB=TOP) want-LIN=1 PASS=kiss by Siti lsquoI want to be kissed by Sitirsquo cprime [eku te=iduk isiq Siti] I PASS=kiss by Siti lsquoI was kissed by Sitirsquo

When the ldquoEqui nominalrdquo of the complement does not bear the Topic relation as in the following examples it cannot occur as a gap

(35) Pancor Ngeno-Ngeneacute a Mele-ng=ku [Siti iduk] (Oslash=A=SUB=NON-TOP) want-LIN=1 Siti Oslashkiss lsquoI want to kiss Sitirsquo aprime [Siti ku=iduk] (PF) Siti 1=Oslashkiss lsquoI kiss Sitirsquo b Mele-ng=ku [Siti ngiduk Oslash] (Oslash=P=OBJ=NON-TOP) want-LIN=1 Siti Nkiss lsquoI want Siti to kiss (me)rsquo


891

bprime [Siti ngiduk eku] (AF) Siti Nkiss I lsquoSiti kisses mersquo

The control pattern is different when the main verb requires a ldquolike-Subjectrdquo in the complement that is coreferential with either the main-clause Subject or Object Here the gap in the complement must hold the Subject and Topic relations simultaneously as shown in the alignment pattern below

(36) Selong Ngeno-Ngeneacute a Ali nyobaq [Oslash ngiduk Siti] (Oslash=A=SUB=TOP) Ali Ntry Nkiss Siti lsquoAli tried to kiss Sitirsquo aprime [Ali ngiduk Siti] (AF) Ali Nkiss Siti lsquoAli kisses Sitirsquo b Ali nyobaq [Oslash te-iduk isiq Siti] (Oslash=P=SUB=TOP) Ali Ntry PASS-kiss by Siti lsquoAli tried to be kissed by Sitirsquo bprime [Ali te-iduk isiq Siti] (Passive) Ali PASS-kiss by Siti lsquoAli was kissed by Sitirsquo

The contrasting form to observe is (37c) below Unlike (34b) above in which the Topic relation is the sole relevant factor for the control phenomenon it is ungrammatical Notice that the P gap of a passive clause is possible as in (36b) above The contrast between the passive complement pattern in (36b) and the PF complement pattern in (37c) cannot be easily explained without recognizing the Subject relation in addition to the Topic relation for the Topic of a passive clause and that of a PF construction would have the identical alignment pattern of ltP=TOPgt without the intermediary SubjectObject relations11

11 Exactly the same PFPassive contrast holds in Standard Indonesian (Bahasa Indonesia) and

Standard Malay (Bahasa Melayu) indicating that PF constructions in these languages are not passive and that the PF topic is not a grammatical Subject contrary to the widely-held belief that they are Observe

Standard IndonesianMalay a AndaAwak men-cium saya (AF) Youyou AF-kiss me lsquoYou kiss mersquo

Masayoshi Shibatani

892

(37) Selong Ngeno-Ngeneacute a Ali nyobaq [na-iduk Siti Oslash] (Oslash=A=SUB=NON-TOP) Ali Ntry 3=Oslashkiss Siti lsquoAli tried to kiss Sitirsquo aprime [na-iduk Siti siq Ali (PF) 3=Oslashkiss Siti by Ali lsquoAli kisses Sitirsquo b Ali nyobaq [Siti ngiduk Oslash] (Oslash=P=OBJ=NON-TOP) Ali Ntry Siti Nkiss (lit) lsquoAli tried Siti to kiss (him)rsquo

bprime [Siti ngiduk Ali] (AF) Siti Nkiss Ali lsquoSiti kisses Alirsquo c Ali nyobaq [Oslash na=iduk isiq Siti] (Oslash=P=OBJ=TOP) Ali Ntry 3=Oslashkiss by Siti (lit) lsquoAli tried Siti to kiss (him)rsquo cprime [Ali na=iduk isiq Siti] (PF) Ali 3=Oslashkiss by Siti lsquoSiti kisses Alirsquo Object-controlled Equi-constructions are similar to the Subject-controlled ones in

that the deleted Equi nominals must hold both Subject and Topic relations

b Saya andaawak cium (PF) I youyou kiss lsquoYou kiss mersquo c Saya di-cium (oleh) Siti (Passive) I PASS-kiss (by) Siti lsquoI was kissed by Sitirsquo aprime AndaAwak coba [Oslash men-cium saya] (Based on AF (a)) youyou try AF-kiss I lsquoYou tried to kiss mersquo bprime Saya coba [Oslash andaawak cium] (Based on PF (b)) I try youyou Oslashkiss cprime Saya coba [Oslash di-cium (oleh) Siti] (Based on Passive (c)) I try PASS-kiss (by) Siti lsquoI tried to be kissed by Sitirsquo


893

(38) Selong Ngeno-Ngeneacute a ku=nyuruq le Siti [Oslash ngiduk Ali] (Oslash=A=SUB=TOP) 1=Norder ART Siti Nkiss Ali lsquoI ordered Siti to kiss Alirsquo aprime [Siti ngiduk Ali] (AF) Siti Nkiss Ali lsquoSiti kissed Alirsquo b Ku=nyuruq Siti [Oslash te-iduq isiq Ali] (Oslash=P=SUB=TOP) 1=Norder Siti PASS-kiss by Ali lsquoI ordered Siti to be kissed by Alirsquo bprime [Siti te-iduq isiq Ali] (Passive) Siti PASSkiss by Ali lsquoSiti was kissed by Alirsquo

If the gap in the complement clause does not bear both Subject and Topic relations

as in the following examples the main-clause Object cannot control it

(39) Selong Ngeno-Ngeneacute a Ku=nyuruq Siti [Ali na=iduk] (Oslash =A=SUB=NON-TOP) 1=Norder Siti Ali 3=Oslashkiss lsquoI ordered Siti to kiss Alirsquo aprime [Ali na=iduk siq Siti] (PF) Ali 3=Oslashkiss by Siti lsquoSiti kisses Sitirsquo

b ku=nyuruq Siti [Oslash na-iduq isiq Ali] (Oslash =P=OBJ=TOP) 1=Norder Siti 3=Oslashkiss by Ali (lit) lsquoI ordered Siti (that) Ali kisses (her)rsquo lsquoI ordered Siti to be kissed by Alirsquo bprime [Siti na=iduq isiq Ali] (PF) Siti 3=Oslashkiss by Ali lsquoAli kissed Sitirsquo

The contrasting pair is the grammatical one containing a gap of a passive comple-

ment whose patient nominal has the alignment of ltP=SUB=TOPgt as in (38b) and the ungrammatical one containing a gap of a PF complement whose patient nominal is aligned in the manner of ltP=OBJ=TOPgt as in (39b) Again without the intermediary Subject and Object relations which are separate from the Topic relation these two nominals would be identical in displaying the alignment pattern of ltP=TOPgt or ltP=SUBgt if the Topic were to be reinterpreted as a Subject An analysis advocating such alignment

Masayoshi Shibatani

894

patterns would not be able to explain the observed contrast in a systematic manner Another phenomenon that requires the SubjectObject relations in addition to the

Topic relation is relativizer selection in Bagu Meno-Meneacute which demands a clear distinction between PF and passive constructions or between an Object-based Topic and a Subject-based Topic If the PF construction were considered to be passive and its Topic were to be regarded as a Subject in Sasak then there would be no distinction between (b) and (c) below as far as the role alignment goes

(40) Bagu Meno-Meneacute A=SUB a Dengan mame=no gitaq dengan nine=ne (AF) person male=that see person female=this lsquoThat man sees this womanrsquo

P=SUB b Dengan nine=ne gitaq=n siq dengan mame=no (PF) person female-this see=3 by person male=that lsquoThat man sees this womanrsquo

P=SUB c Dengan nine=ne te-gitaq siq dengan mame=no (Passive) person female-this PASS-see by person male=that lsquoThis woman is seen by that manrsquo

This analysis would entail complications in accounting for the relativizer selection seen in Bagu Meno-Meneacute where the distinction between the SUB=TOP and the OBJ=TOP alignment plays a crucial rolemdashthe former triggers saq as the basic non-derived SUB=TOP does and the latter the saq-siq relativizer12

(41) Bagu Meno-Meneacute Oslash=A=SUB=TOP a Dengan mame [saq Oslash gitaq dengan nine]=no amaq=k (AF) person male REL see person female=that father=1 lsquoThat man who sees the woman is my fatherrsquo

12 Besides Bagu Meno-Meneacute speakers certain speakers of Ganti Meno-Meneacute make the same

distinction between the Subject relativizer and the Object relativizer


895

Oslash=P=OBJ=TOP b Dengan nine [saq-siq=n gitaq Oslash siq dengan mame]=ne inaq=k (PF) person female REL=3 see by person male=this mother=1 lsquoThis woman whom the man sees is my motherrsquo

Oslash=P=SUB=TOP c Dengan nine [saq Oslash te-gitaq siq dengan mame]=ne inaq=k (Passive)

person female REL PASS-see by person male=this mother-1 lsquoThis woman who is seen by the man is my motherrsquo

Finally describing Sasak relativization in terms of the AH has an unwelcome consequence for the following universal proposed by Comrie amp Keenan (1979661) ldquoAll RC strategies must operate on a continuous segment of the AHrdquo 13

The relevant discussion revolves around the pronoun-retention relativization strategy in Sasak and Sumbawa As widely recognized many languages have more than one relativization strategy such as a gap strategymdashthe one we have been following in this paper so farmdashand a pronoun-retention strategy which leaves a pronominal copy in a relative clause It is also widely acknowledged that the pronoun-retention strategy applies more readily in those positions in the AH where the recovery of the case relation of the gapped element is more difficult namely at the lower end of the AH Comrie amp Keenanrsquos universal above requires that all available relativization strategies operate on a contiguous segment of the AH such that we would not expect a situation in which a pronominal copy is allowed in the OCOMP GEN and SUB positions but not in the OBL IO and DO positions But this is exactly what we find in Sasak and possibly in Sumbawa

As one might expect the Sasak pronoun-retention strategy allows a pronominal copy to appear in the lower end of the AH as the following examples of the relativization on OCOMP and GEN show

(42) Pancor Ngeno-Ngeneacute OCOMP relativized a Ganggasan ante isiq dengan nine ino tall you than person female that lsquoYou are taller than that womanrsquo b dengan nine ino [si ganggasan ante isiq iye] ina-ng=ku person female that REL tall you than her mother-LIN=1 lsquoThat woman whom you are taller than is my motherrsquo

13 Or the generalization stated by Keenan (1985148) in the implicational term as ldquoIf a given

language presents NPrel [a relativized NP] as a pronoun for any position in the Hierarchy then it presents NPrel as a pronoun for all lower positions on the Hierarchyrdquo

Masayoshi Shibatani

896

(43) Bagu Meno-Meneacute GEN relativized a Otak dengan mame=no beleq head person male=that big lsquoThat manrsquos head is bigrsquo b Dengan mame=no [saq otak=n beleq] batur amaq=k person male=that REL head=3 big friend father=1 lsquoThat man whose head is big is my fatherrsquos friendrsquo

The OBL position does not allow a pronominal copy and neither do the two OBJ positions (see (64) below for an ungrammatical form with a pronominal Object clitic)

(44) Pancor Ngeno-Ngeneacute OBL relativized a Oku ngirim surat timpaq batur=kuiye I sent letter to my friendhim lsquoI sent a letter to my friendhimrsquo b Batur [si ngirim=ku surat timpaq iye] pintar friend REL Nsend=1 letter to him smart lsquoThe friend to whom I sent a letter is smartrsquo (45) Bagu Meno-Meneacute OBJ relativized a Beng=k dengan mame=no anak=k give=1 person male=that child=1 lsquoI gave the man my childrsquo b Dengan mame [saq-siq beng=k iye anak=k] batur=k person male REL give=1 him child=1 friend=1 lsquoThe man whom I gave my child is my friendrsquo c Anak=k [saq-siq beng=k dengan mame=no iye] child=1 REL give=1 person male=that him nyengke=n bace buku PROG=3 read book lsquoMy child whom I gave to that man is reading a bookrsquo

Here one might point out that in Sasak only Subjects (our Topics) can be relativized

and argue that the above sentences are ungrammatical anyway because they relativize Objects However his argument does not resolve the problem because we are here looking at a relativization strategy different from the Subjects-only gap strategy and so there is no reason to assume that the Subjects-only constraint applies to the present one


897

the pronoun-retention strategy Also observe the grammatical forms in (42) and (43) where non-Subjects are relativized on

Turning now to the SU position we see that while some Sasak dialects eg Pancor Ngeno-Ngeneacute Narmada Ngeno-Ngeneacute resist retention of pronominal clitics for relativized Subjects there are others that allow them to appear cliticized to auxiliaries

(46) Bagu Meno-Meneacute Dengan mame [saq mele=n gitaq kamu]=no amaq=m person male REL want=3 see you =that father=2 lsquoThe man who wants to see you is your fatherrsquo (47) Darek Meriaq-Mereto Kemu [saq mele=m gitaq keku] iku batur keku you REL want=2 see me that friend my lsquoYou who want to see me is my friendrsquo (48) Pujut Meriaq-Meriku14 Kanak siaq [ke=n bace buku=n] anak=k child this PROG=3 read book=3 child=1 lsquoThis boy who is reading his book is my sonrsquo (49) Sumbawa Jereweh a Aku [adeacute ka=ku pukel bedus] sakit I REL PERF=1 hit goat sick lsquoI who have hit a goat am sickrsquo b Kau [adeacute ka=m pukel bedus] sakit you REL PERF=2 hit goat sick lsquoYou who have hit a goat are sickrsquo c Ali [adeacute ka=ya pukel bedus] sakit Ali REL PERF=3 hit goat sick lsquoAli who has hit a goat is sickrsquo

There are thus several good reasons for not reinterpreting Sasak and Sumbawan

Topics as Subjects and describing their relativization patterns in terms of the Accessibility Hierarchy Indeed analyzing the relativization processes in these languages in terms of Topics apart from the grammatical relations Subject and Object finds an easy answer to the puzzle of why many of these Austronesian languages readily allow GEN to be relativized when they do not allow Objects to be relativized unless they are made Topic What has been analyzed as relativization of GEN in IndonesianMalay and others is in fact not a relativization of the genitive nominal in a possessive construction of the form 14 In this and other dialects of Sasak a relativizer is sometimes optional

Masayoshi Shibatani

898

[John-GEN head]NP for example Rather what is relativized is a Topic which is interpreted as the possessor of the head noun (under the non-derivational analysis) or which is derived from the possessive NP construction (under the derivational analysis) For ease of exposition I shall follow the latter analysis

IndonesianMalay and many other Austronesian languages allow a Topic construction of the following type (50b) where the initial Topic nominal is construed as the possessor of a typically body-part nominal within the following clause

(50) Bagu Meno-Meneacute a [otak dengan mame=no]NP beleq head person male=that big lsquoThat manrsquos head is bigrsquo b Dengan mane=no [otak=n beleq]S person male=that head=3 big (lit) lsquoThat person his head is bigrsquo

Sentence (50a) contains a possessive NP construction as the SubjectTopic of the

sentence and here the meaning is equivalent to what is expressed by the English translation Sentence (50b) on the other hand has a Topic derived from the possessive NP construction outside the basic clause This construction which is difficult to translate into a natural English sentence is similar to the equivalent topic constructions in Japanese and Korean in which the Topics are clearly distinguished from Subjects by special topic markers Notice that the Topic nominal in (50b) above is not a Subject in the ordinary sense It is not lexically predicated by beleq lsquobigrsquomdashthe sentence does not say that the man is big In this type of sentence a clause predicates over the Topic Moreover the Topic here does not cliticize to the verb What cliticizes is the lexically predicated Subject as seen below

(51) Pancor Ngeno-Ngeneacute a [Anak-anak loq Ali] na=pade-sakit child-child ART Ali 3=PL-sick lsquoAlirsquos children are sickrsquo b Loq Ali [anak-anak=na na=pade-sakit] ART Ali child-child=3 3=PL-sick lsquoAli his children are sickrsquo

The relative clause expression of the following type which is usually considered to

be the case of GEN relativization is based on (50b) above not on (50a)15 15 Sneddon (1996288-289) offers a similar analysis based on the topic construction for the


899

(52) Bagu Meno-Meneacute Dengan mame=no [siq [Oslash [otak=n beleq]]] batur amaq=k person male=that REL head=3 big friend father=1 lsquoThe man whose head is big is my fatherrsquos friendrsquo (lit) lsquoThe man who his head is big is my fatherrsquos friendrsquo

This explains why it is so easy for many of these Austronesian languages to relativize

on what looks like the GEN nominal of a possessive NP construction while it is generally difficult to actually relativize such a nominal in other languages as reflected in its low position in the Accessibility Hierarchy There is even more compelling evidence for the Topic-based analysis of the relevant constructions in Bagu Meno-Meneacute Recall that in this language if a Topic is aligned with an Object it triggers the saq-siq relativizer This also happens when a (possessor) Topic is ldquoderivedrdquo from an Object nominal possessive construction and is subsequently relativized

(53) Bagu Meno-Meneacute a Aku empok [otak dengan mame=no]NP

I hit head person male=that lsquoI hit that manrsquos headrsquo b Dengan mame=no [aku empok [otak=n]] person male=that I hit head=3 lsquoThat man I hit his headrsquo c Dengan mame=no [saq-siq=k empok otak=n] batur amaq=k person male=that REL=1 hit head=3 friend father=1 lsquoThat man whose head I hit is my fatherrsquos friendrsquo

While there is a possibility of analyzing a Topic construction like (50b) as a double-

Subject construction or a Topic construction based on a double-Subject construction (see Shibatani 2001) such an analysis is hard to extend to a construction like (53b) because the Topic there ldquoretainsrdquo the Object relation to the verb thereby triggering the saq-siq relativizer16 Notice further that if the GEN nominal of the possessive construction were

Indonesian relative clauses that are normally construed as a relativization of the possessor of a possessive construction (also see Comrie 1989161 Schachter 199396)

16 The nature of the Japanese Topic also differs depending on whether it is Subject-based (eg Taroo=wa Ziroo=o nagutta lsquoTaro=TOP hit Jirorsquo) or Object-based (eg Ziroo=wa Taroo=ga nagutta lsquoJiro=TOP Taro hitrsquo) In the former the Topic nominal possesses all the Subject properties of the nominative Subject eg it triggers Subject honorification whereas in the latter the Topic retains Object properties eg it triggers Object honorification

Masayoshi Shibatani

900

directly relativized as in the analysis found in the literature we would not expect there to be a difference in the relativizer choice for the GEN relation is a relation between the genitive nominal and the head nominal and this relation is constant whether the whole possessive construction functions as a Subject or as an Object

Finally other problematic cases of Sasak relativization find an answer in one Topic construction or another The OCOMP relativization in Pancor Ngeno-Ngeneacute seen earlier and is repeated below in (54c) is based on the Topic construction in (54b) which leaves a pronominal copy in the OCOMP position

(54) Pancor Ngeno-Ngeneacute a Ante ganggasan isiq dengan nine ino you tall than person female that lsquoYou are taller than that womanrsquo b Dengan nine ino [ganggasan ante isiq iye] person female that tall you than she (lit) lsquoThat woman you are taller than herrsquo c Dengan nine ino [si ganggasan ante isiq iye] ina-ng=ku person female that REL tall you than she mother=1 lsquoThe woman whom you are taller than is my motherrsquo

Eades (1998126ff) discusses a number of cases that appear to relativize on oblique

nominals For example the Object of the preposition isiq in (55a) below cannot be relativized (cf earlier example (17b)) Instead we have a form like (55b) where the instrumental phrase occurs before the ldquomainrdquo verb empuk lsquohitrsquo This relative clause expression as correctly noted by Eades comes from the PF construction in (55c) in which the instrumental nominal is a Topic

(55) Puyung Meno-Meneacute a Kanak=no empuk acong isiq tunjang=no child=that hit dog with stick=that lsquoThat child hit the dog with that stickrsquo b Tunjang [saq isiq=n empuk acong isiq kanak]=no leacuteq atas meacuteje stic REL with=3 hit dog by child=that LOC top table lsquoThe stick with which the child hit the dog is on the tablersquo (Eades 1998127) c Tunjang=no isiq=n empuk acong isiq kanak=no stick=that with=3 hit dog by child=that (lit) lsquoThat stick with it the child hit the dogrsquo


901

The PF construction in (55c) above parallels the serial verb PF construction in (56b) below whose AF version is (56a) Isiq in (55c) above however does not seem to behave like a verb in other contexts

(56) Puyung Meno-Meneacute a mu=k kadu tunjang=no empuk acong (AF Eades 1998126) PAST=1 use stick=that hit dog lsquoI use the stick to hit the dogrsquo b tunjang=no kadu=k empuk acong17 (PF Eades 1998126) stick=that use=1 hit dog lsquoI use the stick to hit the dogrsquo

In summary we propose on the basis of our analysis of Sasak (and secondarily on

Sumbawa) that Austronesian AF and PF constructions (and passive constructions in Indonesian languages) have the following argument alignment patterns here illustrated by Standard Indonesian forms paralleling our earlier illustrations of these constructions by Puyung Meno-Meneacute forms in (31) The PF Topic and the passive Topic are clearly distinguished in terms of the grammatical relations Subject and Object with which Patient aligns

(57) Standard Indonesian a AF construction Anda men-cium Saya (A=SU=TOP P=OBJ)18 you N-kiss I lsquoYou kiss mersquo b PF construction Saya anda cium (A=SU P=OBJ=TOP) I you emptykiss lsquoYou kiss mersquo

17 The Pancor Ngeno-Ngeneacute forms paralleling (a) and (b) have the nasal AF form ngadu lsquoto usersquo

and the oral PF form kadu respectively while isiq in this dialect does not show the nasaloral morphological contrast

18 The status of the Object relation is not clear for those languages that have LF and CF constructions that directly select oblique nominals as Topics and for AF constructions in some Philippine and Formosan languages that do not allow a definite patient in them What is important for our consideration is the Subject=Topic alignment such that the Subject=Topic and the Non-Subject =Topic alignment pattern obtain in the AF (and passive) and the PF (and other non-AF) construction respectively throughout the relevant Austronesian languages

Masayoshi Shibatani

902

c Passive construction Saya di-cium oleh Ali (A=OBL P=SU=TOP) I PASS-kiss by Ali lsquoI was kissed by Alirsquo

In the AF construction A holds both Subject and Topic relations while P is aligned

with the Object relation PF and passive constructions differ in terms of both A and P alignments In the PF construction A remains Subject as in the AF construction while the equivalent in the passive construction is an adjunct holding the Oblique relation P in the PF construction on the other hand holds Object and Topic relations whereas the equivalent in the passive construction bears the Subject relation (as in the P nominal of the prototypical passive elsewhere) and the Topic relation We have seen above in terms of the data from eastern Indonesia that while all Topics in (57) share defining syntactic properties (as well as discourse properties such as definiteness) including involvement in relativization they all differ in other respects reflecting differences in the participant roles and the SubjectObject relations that they bear simultaneously

Our methodological stance as discussed in the introduction is that while there are obvious differences in the focus constructions across various Western Malayo-Polynesian and Formosan languages they are considered comparable and can be identified profitably in that all share the crucial properties of the Topic relation that A P and others hold in respective focus constructions In Philippine and Formosan languages as well as Malagasy Topics typically occur sentence finally whereas in Indonesian languages they typically occur sentence initially Philippine and Formosan languages which maintain the V-initial word order have a topic marker (eg Philippine ang) while Indonesian languages with the Topic-initial word order lack such a marker In Ilocano and Puyuma both Topics and Subjects cliticize while in Indonesian languages Topics do not seem to cliticize In Tagalog and Ilocano as well as Puyuma and certain other Formosan languages AF constructions in independent clauses disfavor a definite Patient (with the possibility that they are all syntactically intransitive) while there is no such restriction in Cebuano Atayal and most Indonesian languages Despite these differences the fact that A and P bear the Topic relation in AF and PF constructions respectively across these languages figures importantly in understanding the similarity that they show with respect to various syntactic and discourse phenomena including relativization raising certain types of control construction as well as the definiteness of the Topic nominal and the role it plays in reference tracking

Returning to the issues of relativization a reasonable conclusion to draw from the discussion above is that Sasak and Sumbawan relativizations are best analyzed in terms of the Austronesian Topic relation rather than the grammatical relations Subject and Object and the Accessibility Hierarchy While this conclusion does not invalidate the weaker


903

version of the universal given in the original work by Keenan amp Comrie (1977) cited as (58a) below it does invalidate the sronger versions in Comrie amp Keenan (1979) and Keenan (1985) cited as (58b-c) below

(58) Subject Relative Universal a ldquoAll languages can relativize Subjectsrdquo (Comrie amp Keenan 1979652) b ldquohellip in absolute terms Subjects are the most relativizable of NPrsquosrdquo (Comrie amp Keenan 1979653) c ldquoSubjects are universally the most relativizable of NPsrdquo (Keenan 1985158)

The weaker version of the ldquoSubject Relative Universalrdquo in (58a) is satistifed if some

Subjects in a language are relativized and this is the case in Sasak and Sumbawa since Subject-Topics do relativize However the stronger versions (58b-c) are falsified by these languages In these languages Subjects only relativize when they also bear the Topic relation On the other hand Topics are most readily relativizable of NPrsquos in these languages since they relativize unconditionally

Comrie amp Keenan (1979659) note that ldquoif there turn out to be languages for which the traditional notions of Subject etc are inapplicable then our generalizations make no claim about how relativization should pattern in those languages The generalizations are not of course falsified thereby but their domain of applicability is restrictedrdquo What is significant about the Sasak and Sumbawa data is that ldquothe traditional notions of Subject etcrdquo are applicable to these languages yet they have an additional grammatical relation Topic that is more relativizable than Subject Notice further that if the relativization processes of Western Malayo-Polynesian languages did not operate in terms of Subject and Object which is our general claim there would be quite a large number of languages (perhaps nearly 300) that are not covered by Keenan amp Comriersquos generalization though it is true that these languages all belong to the same language group Formosan languages which are normally classified as Austronesian but non-Malayo-Polyneisan have the same restriction in their relativization processes such that only Topics can be relativized The extent to which other Austronesian languages obey the similar ldquoTopics-only constraintrdquo is not clear at this point The recent study by the author on the Austronesian languages on the islands of Sumbawa Sumba Sabu Palursquoe and Flores classified as Central-Eastern Malayo-Polynesian indicates that they generally maintain the structural AFPF contrast and relativize only Topics despite the fact that they have lost the focus morphology as in many of the dialects in Sasak and Sumbawa What this paper suggests is that all Austronesian languages perhaps numbering over 1200 are worth reconsidering in the light of the Austronesian AFPF structural contrast argued for in this paper For example Samoan antipassive and ergative constructions may be understood as a continuation of the

Masayoshi Shibatani

904

Austronesian AF and PF constructions Perhaps a more revealing understanding of the relativization phenomena in these languages obtains under such a perspective

In addition to Western Malayo-Polynesian languages Keenan amp Comrie (1977) list Aoban (an Austronesian language) and Classical Arabic as Subjects-only languages for the Subject-gap strategy19 They also mention that many European languages have participial RC-forming strategies that apply only to Subjects eg the German form der in seinem Buumlro arbeitende Mann lsquothe man working in his officersquo or its English equivalent I argue elsewhere (Shibatani (forthcoming) and Shibatani Wouk amp Nagaya (forthcoming)) that many relative clauses across languages are actually recycled nominalized clauses which lack an NP that is normally analyzed as a gapped NP in a relative clause The participial clauses in the German example above and its English equivalent are analyzable along the same line Indeed the relevant participial forms have the following kind of usage outside the relative clause context [In seinem Buumlro arbeitend] wurde Peter krank lsquo[Working in his office] Peter got sickrsquo In other words relativizations in many languages including Austronesian languages discussed in this paper and the types of modification constructions exemplified by Keenan amp Comriersquos German example and its English equivalent do not involve the relativization process (gapping of a coreferential NP) that is standardly assumed Instead they simply make use of nominalized clauses found in such expressions as [Working in the office late at night] made Bill very tired in English and Pancor Ngeno-Ngeneacute Beng oku [si leacuteq meacuteje ino] lsquoGive me the one on the tablersquo as modifiers in noun modification constructions like the man [working in the office late at night] and relative clauses like buku[si leacuteq meacuteje ino] lsquothe book which is on the tablersquo In the final section below we shall show that what is suggested here is a more plausible way of approaching the relativization phenomena in Austronesian and many other languages in which relative clauses turn out to be nominalized expressions which function as nominal arguments and nominal predicates outside the relativization context

19 It is not clear whether Classical Arabic has a gap RC strategy since subject relativization involves

a person suffix indexing a relativized subject Examine the following examples provided by Adrian Macelaru

al=rajul-u allaDHii jaa-a DEF ART=man-NOMSG REL comePERF-SUB3MASCSG lsquothe man who camersquo

Object relativization involves a suffixed pronoun as in al=rajul-u allaDHii raay-tu=hu DEF ART=man-NOMSG REL seePERF-SUB1SG=3MASCSG lsquothe man I sawrsquo


905

7 Relativization and nominalization

The final topic dealt with here is no less important than those discussed above in the context of our concern with relativization phenomena as it has bearing on the nature of the gaps in relative clauses and on the broader question of the general typology of relative clauses (Comrie 1998) For expository purposes I have been glossing the particles si siq saq and siq-saq in Sasak dialects and adeacute in Sumbawa as REL as if they were relativizers This labeling is actually a misnomer and a better label for them is NMZ ie nominalizer or nominalization marker as they mark a wide range of elements turned into nominalized forms whose grammatical functions range from simple referring expressions to adverbial subordination Relativization is just one of the functions that nominalized clauses play in these languages as in many other languages of the world (Givoacuten 2001)20 In Sasak as in other Indonesian languages not only clauses but various other kinds of units can also be nominalized including demonstratives ordinal numbers and stative predicates Nominalized expressions behave like regular nominals in functioning both as a Subject and Object and even as a nominal predicate

(59) Puyung Meno-Meneacute a [saq no] baru NMZ that new lsquoThat one is newrsquo b [saq pertame] mame kance [saq kedue] NMZ first male and NMZ second nine (speaking of the children) female lsquoThe first one is male and the second one is femalersquo c Beng=k [saq biru]=no give=1 NMZ blue=that lsquoGive me the blue onersquo (60) Narmada Ngeno-Ngeneacute a Kamu [siq=ku empok] you NMZ=1 hit lsquoYou are the one I hitrsquo b [siq=ku empok] kamu NMZ=1 hit you lsquoThe one I hit is yoursquo

20 A similar point has been made by Englebretson (2008) regarding the Bahasa Indonesian yang

expressions which have often been characterized as relative clauses

Masayoshi Shibatani

906

It is the equational expression of the type seen in (60b) that is often characterized as a cleft construction involving a headless relative clause Analyzing a form like [siq=ku empok] lsquothe one I hitrsquo as a headless relative clause is unmotivated since its function in a cleft construction is to refer to an entity presupposed in the discourse (eg X of the presupposition ldquoI hit Xrdquo) rather than narrowing down the range of possible referents (Keenan 1985142)21

Nominalized clauses in Sasak are used to modify another unit in a number of ways They may be a complement of a noun or a verb as in (61a-b) or modify both concrete and abstract nouns (61a c d e) They also function as an adverbial subordinate clause (61f)

(61) Pancor Ngeno-Ngeneacute a berita [si angku=n loq Ali beruq merariq] news NMZ way=3 ART Ali recently marry lsquothe news that Ali recently got marriedrsquo b Aku lupaq [si angku-n loq Ali wah mbilin kota=no] I forget NMZ way=3 ART Ali PERF leave town=that lsquoI forgot that Ali had left the townrsquo c buku [si ne=bace isiq loq Ali] book NMZ 3=Oslashread by ART Ali lsquothe book that Ali readrsquo d suara [si ne=ngerontok lawang ino isiq loq Ali] sound NMZ 3=knock door the by ART Ali lsquothe sound of Ali knocking on the doorrsquo e waktu [si ku=masih sekolah]=no22 time NMZ 1=still school=that

lsquoAt the time when I was still going to schoolhellipWhen I was going to schoolhelliprsquo

21 Cf Keenan amp Comriersquos (1977) definition of relative clauses Sneddonrsquos (1996300 367) sugges-

tion that these nominalized expressions derive from headed relative clauses via ellipsis of the head nominal is totally unmotivated The actual situation is the other way around namely relative clause expressions obtain when nominalized clauses modify a noun as suggested below

22 Presumably it is this type of noun modification that has given rise to the locative and comitative relativizers of the following type

a Restoran siq taoq Ali mangan mahal gati (Narmada Ngeno-Ngeneacute) restaurant NMZ place Ali eat expensive very lsquoThe restaurant where Ali eats is very expensiversquo

b ta tau dengan ku=gita=sia (Sumbawa Shiohara 200091) this person companion 1SG=lookat=2SG lsquoThis is the person with whom I saw yoursquo


907

f Ali te-semateq [si=ne lekaq leacuteq rurung] Ali PASS-kill NMZ=3 walk in street lsquoAli was killed whenwhile he was walking in the streetrsquo

Among these modification patterns the one containing a gap in the modifying clause as in (61c) is what is normally identified as a relative clause In other words relativization in Sasak as in other Indonesian languages is just one pattern of noun modification involving a nominalized clause as a modifier

Let us now turn to the question of the gap in relative clauses In his 1998 paper Comrie places Japanese relative clauses within the general noun modification pattern of the language as we have suggested above for Sasak Noting the parallel between gapped relative clauses and independent sentences with gaps as in (62) below Comrie following Matsumotorsquos (1988) lead suggests that the relative gap is pronominal anaphor or a zero pronoun rather than a gap created by extraction of the relativized noun

(62) Japanese a kore=ga [Oslash kinoo Oslash katta] hon desu this=NOM yesterday bought book COP lsquoThis is the book that (I) bought Oslash yesterdayrsquo b Oslash kinoo Oslash katta

In Japanese (62b) is a complete and perfect answer to a question such as ldquoHave you bought the book alreadyrdquo Comriersquos position is to assume that it is this elliptic sentence that functions as a modifying clause in the relative clause construction in (62a) While it is true that both full and truncated nominalized clauses occur as modifiers in Sasak and Sumbawa the parallel between relative clauses with a gap and independent sentences with an anaphoric pronoun does not hold in these languages This is because anaphoric pronouns in these languages are typically overt whereas relative clauses as we define them contain a gap For example to the question in (63a) below the appropriate answer in Pancor Ngeno-Ngeneacute would be either (63b) with full pronouns or (63c) with pronominal clitics while the Japanese answer would have gaps for ldquoIrdquo and ldquoitrdquo23

(63) Pancor Ngeno-Ngeneacute a Kumbeq=meq buku=no whatdo=2 book=that lsquoWhat did you do with that bookrsquo

23 See Wouk (2008) for a comprison of anaphoric pronouns between Sasak and Mandarin

Masayoshi Shibatani

908

b Aku nulak-ang ia tipak perpustakaan I return-APPL it to library lsquoI returned it to the libraryrsquo c Ku=nulak-ang=e tipak perpustakaan 1=return-APPL=3 to library lsquoI returned it to the libraryrsquo

The corresponding relative clause however cannot contain the Object clitic

indicating that the relative gap here is an obligatory gap

(64) Pancor Ngeno-Ngeneacute Buku [si ku=tulak-ang=e tipak perpustakaan]=no bagus book NMZ 1=return-APPL=3 to library=that interesting lsquoThe book that I returned to the library was interestingrsquo

Indeed in a simple nominalized form used in the following kind of expression (65) there is no antecedent that can establish an antecedent-anaphor relation with the gap within the nominalized form Also the question of extraction of a relativized item does not arise here because there is nowhere the extracted item can go to

(65) Pancor Ngeno-Ngeneacute Beng oku [si leacuteq meacuteje ino] give I NMZ on table that lsquoGive me the one on that tablersquo

In other words what looks like a relative gap is in fact a gap created by the nominalization process itself

Comrie amp Thompson (1985) recognize two types of lexical nominalization namely the ldquoname of activitystaterdquo type (eg create rarr creation) and the ldquoname of an argumentrdquo type (eg employ rarr employer employ rarr employee) Clausal nominalization also yields these two types of nominal expressions and the one we are interested in here is the one that turns a clause into a nominal form designating the participant of an event Clausal nominalizations in Austronesian languages as in many other languages achieve argument nominalization by deleting the argument designating an event participant Now in Austronesian argument nominalization the focus morphology functions to profile the role of the nominalized argument Thus the Pancor Ngeno-Ngeneacute AF nominalized form si m-bace lsquoNMZ AF-readrsquo profiles the agentive role of the argument namely ldquoone who readsrdquo while the PF counterpart si baca lsquoNMZ PFreadrsquo means ldquowhat is readrdquo This verb-level


909

nominalization can be expanded to include other arguments forming essentially clausal argument nominalization of the following form just as the English nominalized expression reading can be expanded into the reading of the book and into the reading of the book by John or Johnrsquos reading of the book

(66) Pancor Ngeno-Ngeneacute a si Oslash mbace buku ino (AF) NMZ Nread book that lsquothe one who reads the bookrsquo b si Oslash bace=ne isiq loq Ali (PF) NMZ Oslashread=3 by ART Ali lsquothe one read by Alirsquo (67) Bagu Meno-Meneacute a saq Oslash bace buku ino (AF) NMZ read book that lsquothe one who reads the bookrsquo

b saq Oslash n=bace siq Ali (PF) NMZ 3=read by Ali lsquothe one read by Alirsquo

The following Tagalog argument nominalization forms with the three types of focus

affixes show varying degrees of lexicalization

(68) Tagalog a AF mang + dup + base (Schachter amp Otanes 1972103) mam-ba-basa lsquoreaderrsquo man-la-laro lsquoplayerrsquo mang-a-awit lsquosingerrsquo b PF base + -in (Schachter amp Otanes 197299) alaga-in lsquosomething to take care ofrsquo aral-in lsquosomething to studyrsquo awit-in lsquosongrsquo bilih-in lsquosomething to buyrsquo c LF base + -an (Schachter amp Otanes 197298) tagu-an lsquohiding placersquo (tago lsquohidersquo) hiram-an lsquoplace for borrowingrsquo (hiram lsquoborrowrsquo) aklat-an lsquolibraryrsquo (aklat lsquobookrsquo) gulay-an lsquovegetable gardenrsquo (gulay lsquovegetablersquo)

Masayoshi Shibatani

910

The analysis advocated here is reminiscent of what was suggested by Starosta Pawley amp Reid (198283) for the proto-Austronesian focus morphology which was considered to be argument nominalizations that gave rise to the verbal focus system via reanalysis of equational constructions In modern Austronesian languages argument nominalization also appears best analyzed along the line suggested above Thus the gap in an argument nominalization clause necessarily appears to correspond to the Topic nominal in a full clause precisely because the focus morphology indicates the role of the argument profiled and elided through the nominalization process Relativization is no more than modification of a noun by a juxtaposed nominalized clause where the nominalized argument must share the identity with the head noun just as an appositive expression such as ldquoJohn the butcherrdquo requires the identity between the head and the appositive nominal Indeed an entailment relation obtains between the expressions containing the whole appositive construction (69a) and (69b) and those with just appositive modifiers (69aprime) and (69bprime) which can stand on its own

(69) a I saw John the butcher in the market aprime I saw the butcher in the market b Loq Ali mbace buku [si beng=ku iye]=no (Pancor Ngeno-Ngeneacute) ART Ali Nread book NMZ give=1 he=that lsquoAli read the book that I gave himrsquo bprime Loq Ali mbace [si beng=ku iye]=no ART Ali Nread NMZ give=1 he=that lsquoAli read the one I gave himrsquo

The gap in a relative clause in Sasak (and other Indonesian languages including Sumbawa) in other words is due to the nominalization process rather than to the relativization process which simply juxtaposes a head noun and a nominalized clause to form an appositive construction The two related constructions that are often discussed along with relative clauses namely cleft constructions and Wh-questions are equational sentences in which a nominalized clause functions either as a Subject or a predicate as in the following examples It is the involvement of nominalized clauses with a gap profiling the nominalized argument in these constructions that unifies these with relative clause expressions

(70) Pancor Ngeno-Ngeneacute Cleft construction a [Si mbace buku=ne] batur=ku (Based on AF) NMZ Nread book=this friend=1 lsquoThe one who read this book is my friendrsquo


911

b [Si bace=ku] buku=ne (Based on PF) NMZ Oslashread=1 book=this lsquoThe one I read is this bookWhat I read is this bookrsquo Wh-questions c Sai [si mbace buku=ne] (Based on AF) who NMZ Nread book=this lsquoWho is the one that read this bookWho read this bookrsquo d Ape [si bace=meq] (Based on PF) what NMZ Oslashread=2 lsquoWhat is the one that you readWhat did you readrsquo

Again the requirement that the gap in the nominalized clause must match the Subject

and predicate of these constructions is due to the nature of these equational constructions that identify or seek the identity of the nominalized expressions in them Relative clauses cleft constructions and Wh-questions share the property of requiring a specific type of gap in the nominalized clause employed because they all require discourse refential identity between the nominalized clause and the head the predicate or the interrogative pronoun Subject with which the former comes into construction This is a more plausible way of accounting for the gaps in these constructions than the oft-proposed analysis that deletes the Topic nominal under identity with a head noun or that extracts and moves it to another site

8 Conclusion

The nature of the Austronesian Topic nominal and the contrast between AF and PF (and other non-AF) constructions have been highly controversial The situation has been confounded in Philippine and Formosan languages largely because of the challenge they pose in clearly delineating Subjects apart from Topics in them and because most of them do not have passive constructions Many Indonesian languages do have passive constructons but past studies have not scrutinized the possible differences between passive and PF constructions Sasak and Sumbawa are ideal languages in unraveling many of the Austronesian controversies in that they have both robust passive and PF constructions albeit the morphological AFPF contrast is lost in many dialects They also have a clearly delineable Subject cateory that takes in S A and the P of a passive just like the Subject category in English and that contrasts sharply with the Topic category with regard to such phenomena as cliticization and gap-control

Implications from our study of Sasak and Sumbawa are that the Austronesian Topic nominals (ie ang nominals in Tagalog and their equivalents in Malagasy as well as in

Masayoshi Shibatani

912

Formosan languages which are associated with various particles normally glossed as nominative in the Formosan literature) are not really Subjects and that what are like Subjects in these languages are S and A the so-called Actor nominals In this regard the recent formalist literature on Austronesian languages (eg Richards 2000 Pearson 2005) appears correct in analyzing Actor nominals as occupying the Subject positionmdashSPEC of IP and the Topic as occupying a (non-A) specifier position higher in the clause structure

In the traditional framework for analyzing and typologizing relative clauses along the line of Keenan amp Comrie (1977) we have found conclusive evidence that Subjects are not relevant to relativization in Sasak and Sumbawa contrary to the claim by Keenan amp Comrie that the Subjects-only constraint to relativization obtains in many Austronesian languages Their practice of reinterpreting the Austronesian Topic as a Subject and the PF construction as passive was also found unmotivated in view of the fact that a robust Subject category and passive constructions exist independently and that they behave quite differently from the Topic and the PF construction respectively

While the traditional approach to relativization would lead to the analysis in which the Topic instead of the Subject nominal functions as the target of relativization in Sasak and Sumbawa we have argued in the final section that such an analysis is misguided In these and many languages of the world relative clauses are in fact nominalized clauses that function independently outside the relativization context In these languages relativization is simply juxtaposition of a head nominal and a nominalized clause in appositive syntagm Relativization in these languages accordingly involves no gapping or extraction of a coreferential NP in a relative clause as the traditional approach has it the gap in a relative clause is created in the process of argument nominalization This line of argument for the relativization in Austronesian and other languages has been more fully developed in Shibatani Wouk amp Nagaya (forthcoming) where the Subjects-only (or the Topics- only) constraint to Austronesian relativization recognized by Keenan amp Comrie (1977) is an epiphenomenon engendered by the role of focus morphology in argument nominalization and in Austronesian predicate formation


913

References Adelaar Alexander 2005 Malayo-Sumbawan Oceanic Linguistics 442357-388 Arka I Wayan 2003 Balinese Morphosyntax A Lexical-functional Approach Pacific

Linguistics 547 Canberra The Australian National University Austin Peter K (ed) 1998 Sasak Vol 1 Parkville Department of Linguistics and

Applied Linguistics University of Melbourne Austin Peter K 2000a Verbs voice and valence in Sasak Sasak Vol 2 ed by Peter K

Austin 5-23 Parkville Department of Linguistics and Applied Linguistics University of Melbourne

Austin Peter K (ed) 2000b Sasak Vol 2 Parkville Department of Linguistics and Applied Linguistics University of Melbourne

Chung Sandra 1976 On the subject of two passives in Indonesian Subject and Topic ed by Charles N Li 59-99 New York Academic Press

Comrie Bernard 1989 Language Universals and Linguistic Typology Syntax and Morphology (2nd edition) Oxford Basil Blackwell

Comrie Bernard 1998 Rethinking the typology of relative clauses Language Design 1159-86

Comrie Bernard and Kaoru Horie 1995 Complement clauses versus relative clauses some Khmer evidence Discourse Grammar and Typology Papers in Honor of John W M Verhaar ed by Werner Abraham Talmy Givoacuten amp Sandra A Thompson 65-75 Amsterdam amp Philadelphia John Benjamins

Comrie Bernard and Edward L Keenan 1979 Noun phrase accessibility revisited Language 553649-664

Comrie Bernard and Sandra A Thompson 1985 Lexical nominalization Language Typology and Syntactic Description Vol 3 Grammatical Categories and the Lexicon ed by Timothy Shopen 349-398 Cambridge Cambridge University Press

Donohue Mark 1999 A Grammar of Tukang Besi Berlin amp New York Mouton de Gruyter

Durie Mark 1985 A Grammar of Acehnese on the Basis of a Dialect of North Aceh Dordrecht Foris

Eades Yusuf 1998 Relativization Sasak Vol 1 ed by Peter K Austin 119-129 Park-ville Department of Linguistics and Applied Linguistics University of Melbourne

Englebretson Robert 2008 From subordinate clause to noun-phrase yang constructions in colloquial Indonesian Crosslinguistic Studies of Clause Combining The Multi-functionality of Conjunctions ed by Ritva Laury 1-33 Amsterdam amp Philadelphia John Benjamins

Givoacuten Talmy 2001 Syntax An Introduction Vol 2 Amsterdam amp Philadelphia John Benjamins

Masayoshi Shibatani

914

Jacq Pascale 1998 How many dialects are there Sasak Vol 1 ed by Peter K Austin 67-90 Parkville Department of Linguistics and Applied Linguistics University of Melbourne

Keenan Edward L 1985 Relative clauses Language Typology and Syntactic Description Vol 2 Complex Constructions ed by Timothy Shopen 141-170 Cambridge amp New York Cambridge University Press

Keenan Edward L and Bernard Comrie 1977 Noun Phrase accessibility and universal grammar Linguistic Inquiry 8163-99

Klamer Marian 1998 A Grammar of Kambera Berlin amp New York Mouton de Gruyter Kroon Yosep B 1998 The isiq construction and its grammatical relations Sasak Vol 1

ed by Peter K Austin 105-118 Parkville Department of Linguistics and Applied Linguistics University of Melbourne

Matsumoto Yoshiko 1988 Semantic and pragmatics of noun-modifying constructions in Japanese Berkeley Linguistics Society (BLS) 14166-175

Musgrave Simon 2000 A note on animacy hierarchy effects in Sasak and Sumbawa Sasak Vol 2 ed by Peter K Austin 49-56 Parkville Department of Linguistics and Applied Linguistics University of Melbourne

Pearson Matthew 2005 The Malagasy subjecttopic as an Aprime-element Natural Language and Linguistic Theory 232381-457

Richards Norvin 2000 Another look at Tagalog subjects Formal Issues in Austronesian Linguistics ed by Ileana Paul Vivianne Phillips amp Lisa Travis 105-116 Dordrecht amp Boston Kluwer

Ross Malcolm 1995 Reconstructing Proto-Austronesian verbal morphology evidence from Taiwan Austronesian Studies Relating to Taiwan ed by Paul Jen-kuei Li Cheng-hwa Tsang Ying-kuei Huang Dah-an Ho amp Chiu-yu Tseng 727-791 Taipei Institute of History and Philology Academia Sinica

Ross Malcolm 2002 The history and transitivity of Western Austronesian voice and voice-marking The History and Typology of Western Austronesian Voice Systems ed by Fay Wouk amp Malcolm Ross 17-62 Pacific Linguistics 518 Canberra The Australian National University

Schachter Paul 1976 The subject in Philippine languages Topic Actor Actor-Topic of none of the above Subject and Topic ed by Charles N Li 491-518 New York Academic Press

Schachter Paul 199396 The Subject in Tagalog Still None of the Above UCLA Occasional Papers in Linguistics 15 Los Angeles Department of Linguistics University of California

Schachter Paul and Fe T Otanes 1972 Tagalog Reference Grammar Berkeley University of California Press

Shibatani Masayoshi 1991 Grammaticization of topic into subject Approaches to


915

Grammaticalization ed by Elizabeth Closs Traugott amp Bernd Heine 93-133 Amsterdam amp Philadelphia John Benjamins

Shibatani Masayoshi 2001 Non-canonical constructions in Japanese Non-canonical Marking of Subjects and Objects ed by Alexandra Y Aikhenvald Robert M W Dixon amp Masayuki Onishi 307-354 Amsterdam amp Philadelphia John Benjamins

Shibatani Masayoshi (forthcoming) Elements of complex structures where recursion isnrsquot

Shibatani Masayoshi Fay Wouk and Naonori Nagaya (forthcoming) Nominalization and relativization in Austronesian languages

Shiohara Asako 2000 Relativization in Sumbawan Sasak Vol 2 ed by Peter K Austin 85-98 Parkville Department of Linguistics and Applied Linguistics University of Melbourne

Shiohara Asako 2006 Sunbawa-go no Bunpō [A Grammar of Sumbawa] Tokyo Uni-versity of Tokyo dissertation

Sneddon James Neil 1996 Indonesian A Comprehensive Grammar London amp New York Routledge

Starosta Stanley Andrew K Pawley and Lawrence A Reid 198283 The evolution of focus in Austronesian Papers from the Third International Conference on Austrone-sian Linguistics Vol 2 Tracking the Travellers 145-170 Pacific Linguistics C-75 Canberra The Australian National University

Wouk Fay 2002 Voice in the languages of Nusa Tenggara Barat The History and Typology of Western Austronesian Voice Systems ed by Fay Wouk amp Malcolm Ross 285-309 Pacific Linguistics 518 Canberra The Australian National University

Wouk Fay 2008 The syntax of intonation units in Sasak Studies in Language 321 137-162

Wouk Fay and Malcolm Ross (eds) 2002 The History and Typology of Western Austronesian Voice Systems Pacific Linguistics 518 Canberra The Australian National University

[Received 1 May 2008 revised 1 August 2008 accepted 1 September 2008]

Department of Linguistics Rice University 6100 Main Street Houston Texas 77005 USA mattshibatanigmailcom

Masayoshi Shibatani

916

東印尼 Sasak 與 Sumbawa 語的關係化現象

柴谷方良

美國萊斯大學

本文探討東印尼 Sasak 與 Sumbawa 方言中的關係化現象根據研究結

果本文主張 Keenan 與 Comrie 對西馬來亞玻里尼西亞語群的兩項假定並不

成立在西馬來亞玻里尼西亞語群中主題並不能視為主語此外南島語

中的非主事焦點句並不能視為被動句如同印尼語其他西馬來亞玻里尼西

亞語及台灣南島語在 Sasak 及 Sumbawa 方言中主題與主語是兩種並存

的語法關係較之於主語主題更易於關係化此發現推翻 Comrie 與 Keenan所謂「主語最易於關係化」的假設 (Comrie amp Keenan 1979653)此外Sasak與 Sumbawa 方言中的人稱代詞使用機制也推翻了 Comrie 與 Keenan 的另一項

假設亦即「關係化機制必須依循關係化階層 (Accessibility Hierarchy) 中的

順序來運作」(Comrie amp Keenan 1979661)本文主張西馬來亞玻里尼西亞

語中關係結構應視為名物化子句與主要語 (head noun) 的並列與語法關係

並無相干台灣南島語及許多非南島語語言亦呈現類似現象關鍵詞主題主語西馬來亞玻里尼西亞語關係化的不同過程名物化

ltlt ASCII85EncodePages false AllowTransparency false AutoPositionEPSFiles true AutoRotatePages All Binding Left CalGrayProfile (Dot Gain 20) CalRGBProfile (sRGB IEC61966-21) CalCMYKProfile (US Web Coated 050SWOP051 v2) sRGBProfile (sRGB IEC61966-21) CannotEmbedFontPolicy Warning CompatibilityLevel 14 CompressObjects Tags CompressPages true ConvertImagesToIndexed true PassThroughJPEGImages true CreateJDFFile false CreateJobTicket false DefaultRenderingIntent Default DetectBlends true ColorConversionStrategy LeaveColorUnchanged DoThumbnails false EmbedAllFonts true EmbedJobOptions true DSCReportingLevel 0 SyntheticBoldness 100 EmitDSCWarnings false EndPage -1 ImageMemory 1048576 LockDistillerParams false MaxSubsetPct 100 Optimize true OPM 1 ParseDSCComments true ParseDSCCommentsForDocInfo true PreserveCopyPage true PreserveEPSInfo true PreserveHalftoneInfo false PreserveOPIComments false PreserveOverprintSettings true Start SubsetFonts true TransferFunctionInfo Apply UCRandBGInfo Preserve UsePrologue false ColorSettingsFile () AlwaysEmbed [ true ] NeverEmbed [ true ] AntiAliasColorImages false DownsampleColorImages true ColorImageDownsampleType Bicubic ColorImageResolution 300 ColorImageDepth -1 ColorImageDownsampleThreshold 150000 EncodeColorImages true ColorImageFilter DCTEncode AutoFilterColorImages true ColorImageAutoFilterStrategy JPEG ColorACSImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt ColorImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt JPEG2000ColorACSImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt JPEG2000ColorImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt AntiAliasGrayImages false DownsampleGrayImages true GrayImageDownsampleType Bicubic GrayImageResolution 300 GrayImageDepth -1 GrayImageDownsampleThreshold 150000 EncodeGrayImages true GrayImageFilter DCTEncode AutoFilterGrayImages true GrayImageAutoFilterStrategy JPEG GrayACSImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt GrayImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt JPEG2000GrayACSImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt JPEG2000GrayImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt AntiAliasMonoImages false DownsampleMonoImages true MonoImageDownsampleType Bicubic MonoImageResolution 1200 MonoImageDepth -1 MonoImageDownsampleThreshold 150000 EncodeMonoImages true MonoImageFilter CCITTFaxEncode MonoImageDict ltlt K -1 gtgt AllowPSXObjects false PDFX1aCheck false PDFX3Check false PDFXCompliantPDFOnly false PDFXNoTrimBoxError true PDFXTrimBoxToMediaBoxOffset [ 000000 000000 000000 000000 ] PDFXSetBleedBoxToMediaBox true PDFXBleedBoxToTrimBoxOffset [ 000000 000000 000000 000000 ] PDFXOutputIntentProfile () PDFXOutputCondition () PDFXRegistryName (httpwwwcolororg) PDFXTrapped Unknown Description ltlt FRA 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 ENU (Use these settings to create PDF documents with higher image resolution for improved printing quality The PDF documents can be opened with Acrobat and Reader 50 and later) JPN ltFEFF3053306e8a2d5b9a306f30019ad889e350cf5ea6753b50cf3092542b308000200050004400460020658766f830924f5c62103059308b3068304d306b4f7f75283057307e30593002537052376642306e753b8cea3092670059279650306b4fdd306430533068304c3067304d307e305930023053306e8a2d5b9a30674f5c62103057305f00200050004400460020658766f8306f0020004100630072006f0062006100740020304a30883073002000520065006100640065007200200035002e003000204ee5964d30678868793a3067304d307e30593002gt DEU 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 PTB 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 DAN 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 NLD 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 ESP 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 SUO ltFEFF004e00e4006900640065006e002000610073006500740075007300740065006e0020006100760075006c006c006100200076006f0069006400610061006e0020006c0075006f006400610020005000440046002d0061007300690061006b00690072006a006f006a0061002c0020006a006f006900640065006e002000740075006c006f0073007400750073006c00610061007400750020006f006e0020006b006f0072006b006500610020006a00610020006b007500760061006e0020007400610072006b006b007500750073002000730075007500720069002e0020005000440046002d0061007300690061006b00690072006a0061007400200076006f0069006400610061006e0020006100760061007400610020004100630072006f006200610074002d0020006a00610020004100630072006f006200610074002000520065006100640065007200200035002e00300020002d006f0068006a0065006c006d0061006c006c0061002000740061006900200075007500640065006d006d0061006c006c0061002000760065007200730069006f006c006c0061002egt ITA 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 NOR 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 SVE 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 KOR ltFEFFd5a5c0c1b41c0020c778c1c40020d488c9c8c7440020c5bbae300020c704d5740020ace0d574c0c1b3c4c7580020c774bbf8c9c0b97c0020c0acc6a9d558c5ec00200050004400460020bb38c11cb97c0020b9ccb4e4b824ba740020c7740020c124c815c7440020c0acc6a9d558c2edc2dcc624002e0020c7740020c124c815c7440020c0acc6a9d558c5ec0020b9ccb4e000200050004400460020bb38c11cb2940020004100630072006f0062006100740020bc0f002000520065006100640065007200200035002e00300020c774c0c1c5d0c11c0020c5f40020c2180020c788c2b5b2c8b2e4002egt CHS ltFEFF4f7f75288fd94e9b8bbe7f6e521b5efa76840020005000440046002065876863ff0c5c065305542b66f49ad8768456fe50cf52068fa87387ff0c4ee563d09ad8625353708d2891cf30028be5002000500044004600206587686353ef4ee54f7f752800200020004100630072006f00620061007400204e0e002000520065006100640065007200200035002e00300020548c66f49ad87248672c62535f003002gt CHT ltFEFF4f7f752890194e9b8a2d5b9a5efa7acb76840020005000440046002065874ef65305542b8f039ad876845f7150cf89e367905ea6ff0c4fbf65bc63d066075217537054c18cea3002005000440046002065874ef653ef4ee54f7f75280020004100630072006f0062006100740020548c002000520065006100640065007200200035002e0030002053ca66f465b07248672c4f86958b555f3002gt gtgtgtgt setdistillerparamsltlt HWResolution [2400 2400] PageSize [612000 792000]gtgt setpagedevice

Masayoshi Shibatani

866

in establishing the higher end of the hierarchy of grammatical relations known as the Accessibility Hierarchy Philippine-type languages are characterized in terms of what is known as the focus morphology that marks in the verb the grammatical role of a nominal element selected as the primary argument of a clause With the crucial assumption that this primary argument is Subject in these languages Keenan amp Comrie observe that these languages obey the ldquoSubjects-onlyrdquo constraint representing languages that relativize only on Subjects and draw the universal ldquoin absolute terms Subjects are the most relativizable of NPrsquosrdquo (Comrie amp Keenan 1979653)

Philippine-type languages where the more faithful reflexes of the Proto Austronesian (PAn) focus opposition are seen are found in Taiwan and Madagascar besides the Philippines The PAn focus system is generally believed to have had three or possibly four categories of AF (Actor-focus) PF (Patient-focus) LF (Location-focus) and IF (Instrumental-focus) eg the indicative neutral series Actor ltumgt Patient -ən Location -an InstrumentalCircumstantial Si- (cf Ross 200233) The four-way PAn focus morphological opposition has been reduced to a three- or two-way system in many of the languages of Indonesia (and elsewhere) where the two-way morphological opposition is seen in Standard Indonesian (Bahasa Indonesia) Javanese Balinese and others in terms of a nasal prefix (for the AF) and a zero prefix or oral form (for the PF) Keenan amp Comriersquos Subjects-only constraint on relativization which Ross (1995730) believes to have obtained in PAn is clearly seen in those languages with the two-way morphological focus opposition The Actor nominal in an AF construction with a nasal verbal prefix and the Patient nominal in a PF construction with a zero prefixmdashboth of which Keenan amp Comrie would consider to be Subjectsmdash are the only directly relativizable arguments in these languages The two-way morphological focus opposition seen in major Indonesian languages as well as in Standard Malay (Bahasa Melayu) is being gradually lost in Sasak and Sumbawa in eastern Indonesia It is an interesting question whether the Subjects-only constraint on relativization has also been lost in the languages that have lost the morpho-logical focus contrast as in many dialects of Sasak and Sumbawa and this is one of the questions addressed in this paper

The paper consists of a descriptive and a theoretical component It offers a detailed description of the relativization phenomena seen in various Sasak dialects and primarily two Sumbawan dialects While typological studies such as Keenan amp Comrie (1977) and others by necessity heavily reply on descriptive grammars of various languages it is true

were almost exclusively based This work will be critically reviewed below My special thanks

go to Nur Ahmadi for his assistance in gathering and interpreting the relevant data I am indebted to Fay Wouk and I Wayan Arka Co-PIs of the NSF project mentioned above who have provided me with comments very helpful in improving this paper I also thank Naonori Nagaya for expertly drawing the three maps in this article


867






Masayoshi Shibatani

868






869

2 Sasak and Sumbawa





Masayoshi Shibatani

870





871





Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學











ltlt ASCII85EncodePages false AllowTransparency false AutoPositionEPSFiles true AutoRotatePages All Binding Left CalGrayProfile (Dot Gain 20) CalRGBProfile (sRGB IEC61966-21) CalCMYKProfile (US Web Coated 050SWOP051 v2) sRGBProfile (sRGB IEC61966-21) CannotEmbedFontPolicy Warning CompatibilityLevel 14 CompressObjects Tags CompressPages true ConvertImagesToIndexed true PassThroughJPEGImages true CreateJDFFile false CreateJobTicket false DefaultRenderingIntent Default DetectBlends true ColorConversionStrategy LeaveColorUnchanged DoThumbnails false EmbedAllFonts true EmbedJobOptions true DSCReportingLevel 0 SyntheticBoldness 100 EmitDSCWarnings false EndPage -1 ImageMemory 1048576 LockDistillerParams false MaxSubsetPct 100 Optimize true OPM 1 ParseDSCComments true ParseDSCCommentsForDocInfo true PreserveCopyPage true PreserveEPSInfo true PreserveHalftoneInfo false PreserveOPIComments false PreserveOverprintSettings true StartPage 1 SubsetFonts true TransferFunctionInfo Apply UCRandBGInfo Preserve UsePrologue false ColorSettingsFile () AlwaysEmbed [ true ] NeverEmbed [ true ] AntiAliasColorImages false DownsampleColorImages true ColorImageDownsampleType Bicubic ColorImageResolution 300 ColorImageDepth -1 ColorImageDownsampleThreshold 150000 EncodeColorImages true ColorImageFilter DCTEncode AutoFilterColorImages true ColorImageAutoFilterStrategy JPEG ColorACSImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt ColorImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt JPEG2000ColorACSImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt JPEG2000ColorImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt AntiAliasGrayImages false DownsampleGrayImages true GrayImageDownsampleType Bicubic GrayImageResolution 300 GrayImageDepth -1 GrayImageDownsampleThreshold 150000 EncodeGrayImages true GrayImageFilter DCTEncode AutoFilterGrayImages true GrayImageAutoFilterStrategy JPEG GrayACSImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt GrayImageDict ltlt QFactor 015 HSamples [1 1 1 1] VSamples [1 1 1 1] gtgt JPEG2000GrayACSImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt JPEG2000GrayImageDict ltlt TileWidth 256 TileHeight 256 Quality 30 gtgt AntiAliasMonoImages false DownsampleMonoImages true MonoImageDownsampleType Bicubic MonoImageResolution 1200 MonoImageDepth -1 MonoImageDownsampleThreshold 150000 EncodeMonoImages true MonoImageFilter CCITTFaxEncode MonoImageDict ltlt K -1 gtgt AllowPSXObjects false PDFX1aCheck false PDFX3Check false PDFXCompliantPDFOnly false PDFXNoTrimBoxError true PDFXTrimBoxToMediaBoxOffset [ 000000 000000 000000 000000 ] PDFXSetBleedBoxToMediaBox true PDFXBleedBoxToTrimBoxOffset [ 000000 000000 000000 000000 ] PDFXOutputIntentProfile () PDFXOutputCondition () PDFXRegistryName (httpwwwcolororg) PDFXTrapped Unknown Description ltlt FRA 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 ENU (Use these settings to create PDF documents with higher image resolution for improved printing quality The PDF documents can be opened with Acrobat and Reader 50 and later) JPN ltFEFF3053306e8a2d5b9a306f30019ad889e350cf5ea6753b50cf3092542b308000200050004400460020658766f830924f5c62103059308b3068304d306b4f7f75283057307e30593002537052376642306e753b8cea3092670059279650306b4fdd306430533068304c3067304d307e305930023053306e8a2d5b9a30674f5c62103057305f00200050004400460020658766f8306f0020004100630072006f0062006100740020304a30883073002000520065006100640065007200200035002e003000204ee5964d30678868793a3067304d307e30593002gt DEU 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 PTB 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 DAN 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 NLD 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 ESP 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 SUO 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 ITA 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 NOR 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 SVE ltFEFF0041006e007600e4006e00640020006400650020006800e4007200200069006e0073007400e4006c006c006e0069006e006700610072006e00610020006e00e40072002000640075002000760069006c006c00200073006b0061007000610020005000440046002d0064006f006b0075006d0065006e00740020006d006500640020006800f6006700720065002000620069006c0064007500700070006c00f60073006e0069006e00670020006f006300680020006400e40072006d006500640020006600e50020006200e400740074007200650020007500740073006b00720069006600740073006b00760061006c0069007400650074002e0020005000440046002d0064006f006b0075006d0065006e00740065006e0020006b0061006e002000f600700070006e006100730020006d006500640020004100630072006f0062006100740020006f00630068002000520065006100640065007200200035002e003000200065006c006c00650072002000730065006e006100720065002egt KOR ltFEFFd5a5c0c1b41c0020c778c1c40020d488c9c8c7440020c5bbae300020c704d5740020ace0d574c0c1b3c4c7580020c774bbf8c9c0b97c0020c0acc6a9d558c5ec00200050004400460020bb38c11cb97c0020b9ccb4e4b824ba740020c7740020c124c815c7440020c0acc6a9d558c2edc2dcc624002e0020c7740020c124c815c7440020c0acc6a9d558c5ec0020b9ccb4e000200050004400460020bb38c11cb2940020004100630072006f0062006100740020bc0f002000520065006100640065007200200035002e00300020c774c0c1c5d0c11c0020c5f40020c2180020c788c2b5b2c8b2e4002egt CHS ltFEFF4f7f75288fd94e9b8bbe7f6e521b5efa76840020005000440046002065876863ff0c5c065305542b66f49ad8768456fe50cf52068fa87387ff0c4ee563d09ad8625353708d2891cf30028be5002000500044004600206587686353ef4ee54f7f752800200020004100630072006f00620061007400204e0e002000520065006100640065007200200035002e00300020548c66f49ad87248672c62535f003002gt CHT ltFEFF4f7f752890194e9b8a2d5b9a5efa7acb76840020005000440046002065874ef65305542b8f039ad876845f7150cf89e367905ea6ff0c4fbf65bc63d066075217537054c18cea3002005000440046002065874ef653ef4ee54f7f75280020004100630072006f0062006100740020548c002000520065006100640065007200200035002e0030002053ca66f465b07248672c4f86958b555f3002gt gtgtgtgt setdistillerparamsltlt HWResolution [2400 2400] PageSize [612000 792000]gtgt setpagedevice


867






Masayoshi Shibatani

868






869

2 Sasak and Sumbawa





Masayoshi Shibatani

870





871





Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

868






869

2 Sasak and Sumbawa





Masayoshi Shibatani

870





871





Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













869

2 Sasak and Sumbawa





Masayoshi Shibatani

870





871





Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

870





871





Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













871





Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

872



873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













873






Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

874







875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













875




Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

876






877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













877








Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

878


4 Relativization








879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













879







Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

880









881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













881







Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

882






883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













883




Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

884









885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













885











Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

886





887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













887




Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

888


6 Subject or Topic





889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













889







Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

890






891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













891








Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

892






893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













893








Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

894











895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













895










Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

896







897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













897






Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

898










899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













899










Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

900








901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













901









Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

902







903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













903






Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

904








905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













905






Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

906











907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













907








Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

908










909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













909







Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

910






911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













911




8 Conclusion



Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

912





913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













913

















Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

914

















915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學













915














Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












Masayoshi Shibatani

916


柴谷方良

美國萊斯大學












relativization in sasak and sumbawa, eastern indonesia*

Documents